idnits 2.17.1 draft-ietf-rtcweb-jsep-07.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. 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 document has examples using IPv4 documentation addresses according to RFC6890, but does not use any IPv6 documentation addresses. Maybe there should be IPv6 examples, too? Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (July 4, 2014) is 3585 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in RFCs) == Unused Reference: 'RFC5124' is defined on line 1719, but no explicit reference was found in the text == Outdated reference: A later version (-17) exists of draft-ietf-mmusic-msid-01 == Outdated reference: A later version (-26) exists of draft-ietf-mmusic-sctp-sdp-04 == Outdated reference: A later version (-54) exists of draft-ietf-mmusic-sdp-bundle-negotiation-04 == Outdated reference: A later version (-19) exists of draft-ietf-mmusic-sdp-mux-attributes-01 == Outdated reference: A later version (-11) exists of draft-ietf-rtcweb-audio-02 == Outdated reference: A later version (-09) exists of draft-ietf-rtcweb-data-protocol-04 == Outdated reference: A later version (-26) exists of draft-ietf-rtcweb-rtp-usage-09 == Outdated reference: A later version (-12) exists of draft-ietf-rtcweb-security-06 == Outdated reference: A later version (-20) exists of draft-ietf-rtcweb-security-arch-09 ** Obsolete normative reference: RFC 4566 (Obsoleted by RFC 8866) ** Obsolete normative reference: RFC 4572 (Obsoleted by RFC 8122) ** Obsolete normative reference: RFC 5245 (Obsoleted by RFC 8445, RFC 8839) ** Obsolete normative reference: RFC 5285 (Obsoleted by RFC 8285) == Outdated reference: A later version (-02) exists of draft-ietf-mmusic-trickle-ice-00 == Outdated reference: A later version (-08) exists of draft-nandakumar-rtcweb-sdp-02 Summary: 4 errors (**), 0 flaws (~~), 13 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Uberti 3 Internet-Draft Google 4 Intended status: Standards Track C. Jennings 5 Expires: January 5, 2015 Cisco 6 E. Rescorla, Ed. 7 Mozilla 8 July 4, 2014 10 Javascript Session Establishment Protocol 11 draft-ietf-rtcweb-jsep-07 13 Abstract 15 This document describes the mechanisms for allowing a Javascript 16 application to control the signaling plane of a multimedia session 17 via the interface specified in the W3C RTCPeerConnection API, and 18 discusses how this relates to existing signaling protocols. 20 Status of this Memo 22 This Internet-Draft is submitted in full conformance with the 23 provisions of BCP 78 and BCP 79. 25 Internet-Drafts are working documents of the Internet Engineering 26 Task Force (IETF). Note that other groups may also distribute 27 working documents as Internet-Drafts. The list of current Internet- 28 Drafts is at http://datatracker.ietf.org/drafts/current/. 30 Internet-Drafts are draft documents valid for a maximum of six months 31 and may be updated, replaced, or obsoleted by other documents at any 32 time. It is inappropriate to use Internet-Drafts as reference 33 material or to cite them other than as "work in progress." 35 This Internet-Draft will expire on January 5, 2015. 37 Copyright Notice 39 Copyright (c) 2014 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents 44 (http://trustee.ietf.org/license-info) in effect on the date of 45 publication of this document. Please review these documents 46 carefully, as they describe your rights and restrictions with respect 47 to this document. Code Components extracted from this document must 48 include Simplified BSD License text as described in Section 4.e of 49 the Trust Legal Provisions and are provided without warranty as 50 described in the Simplified BSD License. 52 Table of Contents 54 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 55 1.1. General Design of JSEP . . . . . . . . . . . . . . . . . . 4 56 1.2. Other Approaches Considered . . . . . . . . . . . . . . . 5 57 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 6 58 3. Semantics and Syntax . . . . . . . . . . . . . . . . . . . . . 6 59 3.1. Signaling Model . . . . . . . . . . . . . . . . . . . . . 6 60 3.2. Session Descriptions and State Machine . . . . . . . . . . 7 61 3.3. Session Description Format . . . . . . . . . . . . . . . . 9 62 3.4. ICE . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 63 3.4.1. ICE Candidate Trickling . . . . . . . . . . . . . . . 10 64 3.4.1.1. ICE Candidate Format . . . . . . . . . . . . . . . 11 65 3.4.2. ICE Candidate Pool . . . . . . . . . . . . . . . . . . 11 66 3.5. Interactions With Forking . . . . . . . . . . . . . . . . 12 67 3.5.1. Sequential Forking . . . . . . . . . . . . . . . . . . 12 68 3.5.2. Parallel Forking . . . . . . . . . . . . . . . . . . . 13 69 4. Interface . . . . . . . . . . . . . . . . . . . . . . . . . . 13 70 4.1. Methods . . . . . . . . . . . . . . . . . . . . . . . . . 14 71 4.1.1. Constructor . . . . . . . . . . . . . . . . . . . . . 14 72 4.1.2. createOffer . . . . . . . . . . . . . . . . . . . . . 14 73 4.1.3. createAnswer . . . . . . . . . . . . . . . . . . . . . 15 74 4.1.4. SessionDescriptionType . . . . . . . . . . . . . . . . 16 75 4.1.4.1. Use of Provisional Answers . . . . . . . . . . . . 17 76 4.1.4.2. Rollback . . . . . . . . . . . . . . . . . . . . . 18 77 4.1.5. setLocalDescription . . . . . . . . . . . . . . . . . 18 78 4.1.6. setRemoteDescription . . . . . . . . . . . . . . . . . 19 79 4.1.7. localDescription . . . . . . . . . . . . . . . . . . . 19 80 4.1.8. remoteDescription . . . . . . . . . . . . . . . . . . 20 81 4.1.9. updateIce . . . . . . . . . . . . . . . . . . . . . . 20 82 4.1.10. addIceCandidate . . . . . . . . . . . . . . . . . . . 20 83 5. SDP Interaction Procedures . . . . . . . . . . . . . . . . . . 20 84 5.1. Requirements Overview . . . . . . . . . . . . . . . . . . 21 85 5.1.1. Implementation Requirements . . . . . . . . . . . . . 21 86 5.1.2. Usage Requirements . . . . . . . . . . . . . . . . . . 22 87 5.2. Constructing an Offer . . . . . . . . . . . . . . . . . . 22 88 5.2.1. Initial Offers . . . . . . . . . . . . . . . . . . . . 22 89 5.2.2. Subsequent Offers . . . . . . . . . . . . . . . . . . 26 90 5.2.3. Options Handling . . . . . . . . . . . . . . . . . . . 28 91 5.2.3.1. OfferToReceiveAudio . . . . . . . . . . . . . . . 28 92 5.2.3.2. OfferToReceiveVideo . . . . . . . . . . . . . . . 29 93 5.2.3.3. VoiceActivityDetection . . . . . . . . . . . . . . 29 94 5.2.3.4. IceRestart . . . . . . . . . . . . . . . . . . . . 29 95 5.3. Generating an Answer . . . . . . . . . . . . . . . . . . . 29 96 5.3.1. Initial Answers . . . . . . . . . . . . . . . . . . . 30 97 5.3.2. Subsequent Answers . . . . . . . . . . . . . . . . . . 33 98 5.3.3. Options Handling . . . . . . . . . . . . . . . . . . . 33 99 5.4. Parsing an Offer . . . . . . . . . . . . . . . . . . . . . 33 100 5.5. Parsing an Answer . . . . . . . . . . . . . . . . . . . . 33 101 5.6. Applying a Local Description . . . . . . . . . . . . . . . 33 102 5.7. Applying a Remote Description . . . . . . . . . . . . . . 33 103 6. Configurable SDP Parameters . . . . . . . . . . . . . . . . . 33 104 7. Security Considerations . . . . . . . . . . . . . . . . . . . 34 105 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 35 106 9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 35 107 10. References . . . . . . . . . . . . . . . . . . . . . . . . . . 35 108 10.1. Normative References . . . . . . . . . . . . . . . . . . . 35 109 10.2. Informative References . . . . . . . . . . . . . . . . . . 38 110 Appendix A. JSEP Implementation Examples . . . . . . . . . . . . 39 111 A.1. Example API Flows . . . . . . . . . . . . . . . . . . . . 39 112 A.1.1. Call using ROAP . . . . . . . . . . . . . . . . . . . 39 113 A.1.2. Call using XMPP . . . . . . . . . . . . . . . . . . . 40 114 A.1.3. Adding video to a call, using XMPP . . . . . . . . . . 42 115 A.1.4. Simultaneous add of video streams, using XMPP . . . . 42 116 A.1.5. Call using SIP . . . . . . . . . . . . . . . . . . . . 43 117 A.1.6. Handling early media (e.g. 1-800-GO FEDEX), using 118 SIP . . . . . . . . . . . . . . . . . . . . . . . . . 44 119 A.2. Example Session Descriptions . . . . . . . . . . . . . . . 45 120 A.2.1. createOffer . . . . . . . . . . . . . . . . . . . . . 45 121 A.2.2. createAnswer . . . . . . . . . . . . . . . . . . . . . 47 122 A.2.3. Call Flows . . . . . . . . . . . . . . . . . . . . . . 49 123 Appendix B. Change log . . . . . . . . . . . . . . . . . . . . . 49 124 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 50 126 1. Introduction 128 This document describes how the W3C WEBRTC RTCPeerConnection 129 interface[W3C.WD-webrtc-20140617] is used to control the setup, 130 management and teardown of a multimedia session. 132 1.1. General Design of JSEP 134 The thinking behind WebRTC call setup has been to fully specify and 135 control the media plane, but to leave the signaling plane up to the 136 application as much as possible. The rationale is that different 137 applications may prefer to use different protocols, such as the 138 existing SIP or Jingle call signaling protocols, or something custom 139 to the particular application, perhaps for a novel use case. In this 140 approach, the key information that needs to be exchanged is the 141 multimedia session description, which specifies the necessary 142 transport and media configuration information necessary to establish 143 the media plane. 145 With these considerations in mind, this document describes the 146 Javascript Session Establishment Protocol (JSEP) that allows for full 147 control of the signaling state machine from Javascript. JSEP removes 148 the browser almost entirely from the core signaling flow, which is 149 instead handled by the Javascript making use of two interfaces: (1) 150 passing in local and remote session descriptions and (2) interacting 151 with the ICE state machine. 153 In this document, the use of JSEP is described as if it always occurs 154 between two browsers. Note though in many cases it will actually be 155 between a browser and some kind of server, such as a gateway or MCU. 156 This distinction is invisible to the browser; it just follows the 157 instructions it is given via the API. 159 JSEP's handling of session descriptions is simple and 160 straightforward. Whenever an offer/answer exchange is needed, the 161 initiating side creates an offer by calling a createOffer() API. The 162 application optionally modifies that offer, and then uses it to set 163 up its local config via the setLocalDescription() API. The offer is 164 then sent off to the remote side over its preferred signaling 165 mechanism (e.g., WebSockets); upon receipt of that offer, the remote 166 party installs it using the setRemoteDescription() API. 168 When the call is accepted, the callee uses the createAnswer() API to 169 generate an appropriate answer, applies it using 170 setLocalDescription(), and sends the answer back to the initiator 171 over the signaling channel. When the offerer gets that answer, it 172 installs it using setRemoteDescription(), and initial setup is 173 complete. This process can be repeated for additional offer/answer 174 exchanges. 176 Regarding ICE [RFC5245], JSEP decouples the ICE state machine from 177 the overall signaling state machine, as the ICE state machine must 178 remain in the browser, because only the browser has the necessary 179 knowledge of candidates and other transport info. Performing this 180 separation also provides additional flexibility; in protocols that 181 decouple session descriptions from transport, such as Jingle, the 182 session description can be sent immediately and the transport 183 information can be sent when available. In protocols that don't, 184 such as SIP, the information can be used in the aggregated form. 185 Sending transport information separately can allow for faster ICE and 186 DTLS startup, since ICE checks can start as soon as any transport 187 information is available rather than waiting for all of it. 189 Through its abstraction of signaling, the JSEP approach does require 190 the application to be aware of the signaling process. While the 191 application does not need to understand the contents of session 192 descriptions to set up a call, the application must call the right 193 APIs at the right times, convert the session descriptions and ICE 194 information into the defined messages of its chosen signaling 195 protocol, and perform the reverse conversion on the messages it 196 receives from the other side. 198 One way to mitigate this is to provide a Javascript library that 199 hides this complexity from the developer; said library would 200 implement a given signaling protocol along with its state machine and 201 serialization code, presenting a higher level call-oriented interface 202 to the application developer. For example, libraries exist to adapt 203 the JSEP API into an API suitable for a SIP or XMPP. Thus, JSEP 204 provides greater control for the experienced developer without 205 forcing any additional complexity on the novice developer. 207 1.2. Other Approaches Considered 209 One approach that was considered instead of JSEP was to include a 210 lightweight signaling protocol. Instead of providing session 211 descriptions to the API, the API would produce and consume messages 212 from this protocol. While providing a more high-level API, this put 213 more control of signaling within the browser, forcing the browser to 214 have to understand and handle concepts like signaling glare. In 215 addition, it prevented the application from driving the state machine 216 to a desired state, as is needed in the page reload case. 218 A second approach that was considered but not chosen was to decouple 219 the management of the media control objects from session 220 descriptions, instead offering APIs that would control each component 221 directly. This was rejected based on a feeling that requiring 222 exposure of this level of complexity to the application programmer 223 would not be beneficial; it would result in an API where even a 224 simple example would require a significant amount of code to 225 orchestrate all the needed interactions, as well as creating a large 226 API surface that needed to be agreed upon and documented. In 227 addition, these API points could be called in any order, resulting in 228 a more complex set of interactions with the media subsystem than the 229 JSEP approach, which specifies how session descriptions are to be 230 evaluated and applied. 232 One variation on JSEP that was considered was to keep the basic 233 session description-oriented API, but to move the mechanism for 234 generating offers and answers out of the browser. Instead of 235 providing createOffer/createAnswer methods within the browser, this 236 approach would instead expose a getCapabilities API which would 237 provide the application with the information it needed in order to 238 generate its own session descriptions. This increases the amount of 239 work that the application needs to do; it needs to know how to 240 generate session descriptions from capabilities, and especially how 241 to generate the correct answer from an arbitrary offer and the 242 supported capabilities. While this could certainly be addressed by 243 using a library like the one mentioned above, it basically forces the 244 use of said library even for a simple example. Providing 245 createOffer/createAnswer avoids this problem, but still allows 246 applications to generate their own offers/answers (to a large extent) 247 if they choose, using the description generated by createOffer as an 248 indication of the browser's capabilities. 250 2. Terminology 252 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 253 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 254 document are to be interpreted as described in [RFC2119]. 256 3. Semantics and Syntax 258 3.1. Signaling Model 260 JSEP does not specify a particular signaling model or state machine, 261 other than the generic need to exchange SDP media descriptions in the 262 fashion described by [RFC3264] (offer/answer) in order for both sides 263 of the session to know how to conduct the session. JSEP provides 264 mechanisms to create offers and answers, as well as to apply them to 265 a session. However, the browser is totally decoupled from the actual 266 mechanism by which these offers and answers are communicated to the 267 remote side, including addressing, retransmission, forking, and glare 268 handling. These issues are left entirely up to the application; the 269 application has complete control over which offers and answers get 270 handed to the browser, and when. 272 +-----------+ +-----------+ 273 | Web App |<--- App-Specific Signaling -->| Web App | 274 +-----------+ +-----------+ 275 ^ ^ 276 | SDP | SDP 277 V V 278 +-----------+ +-----------+ 279 | Browser |<----------- Media ------------>| Browser | 280 +-----------+ +-----------+ 282 Figure 1: JSEP Signaling Model 284 3.2. Session Descriptions and State Machine 286 In order to establish the media plane, the user agent needs specific 287 parameters to indicate what to transmit to the remote side, as well 288 as how to handle the media that is received. These parameters are 289 determined by the exchange of session descriptions in offers and 290 answers, and there are certain details to this process that must be 291 handled in the JSEP APIs. 293 Whether a session description applies to the local side or the remote 294 side affects the meaning of that description. For example, the list 295 of codecs sent to a remote party indicates what the local side is 296 willing to receive, which, when intersected with the set of codecs 297 the remote side supports, specifies what the remote side should send. 298 However, not all parameters follow this rule; for example, the SRTP 299 parameters [RFC4568] sent to a remote party indicate what the local 300 side will use to encrypt, and thereby what the remote party should 301 expect to receive; the remote party will have to accept these 302 parameters, with no option to choose a different value. [[OPEN 303 ISSUE: This is not correct because we removed SDES 304 (https://github.com/rtcweb-wg/jsep/issues/10)]] 306 In addition, various RFCs put different conditions on the format of 307 offers versus answers. For example, a offer may propose multiple 308 SRTP configurations, but an answer may only contain a single SRTP 309 configuration. [[OPEN ISSUE: See issue 10 above.]] 311 Lastly, while the exact media parameters are only known only after an 312 offer and an answer have been exchanged, it is possible for the 313 offerer to receive media after they have sent an offer and before 314 they have received an answer. To properly process incoming media in 315 this case, the offerer's media handler must be aware of the details 316 of the offer before the answer arrives. 318 Therefore, in order to handle session descriptions properly, the user 319 agent needs: 320 1. To know if a session description pertains to the local or remote 321 side. 322 2. To know if a session description is an offer or an answer. 323 3. To allow the offer to be specified independently of the answer. 324 JSEP addresses this by adding both setLocalDescription and 325 setRemoteDescription methods and having session description objects 326 contain a type field indicating the type of session description being 327 supplied. This satisfies the requirements listed above for both the 328 offerer, who first calls setLocalDescription(sdp [offer]) and then 329 later setRemoteDescription(sdp [answer]), as well as for the 330 answerer, who first calls setRemoteDescription(sdp [offer]) and then 331 later setLocalDescription(sdp [answer]). 333 JSEP also allows for an answer to be treated as provisional by the 334 application. Provisional answers provide a way for an answerer to 335 communicate initial session parameters back to the offerer, in order 336 to allow the session to begin, while allowing a final answer to be 337 specified later. This concept of a final answer is important to the 338 offer/answer model; when such an answer is received, any extra 339 resources allocated by the caller can be released, now that the exact 340 session configuration is known. These "resources" can include things 341 like extra ICE components, TURN candidates, or video decoders. 342 Provisional answers, on the other hand, do no such deallocation 343 results; as a result, multiple dissimilar provisional answers can be 344 received and applied during call setup. 346 In [RFC3264], the constraint at the signaling level is that only one 347 offer can be outstanding for a given session, but at the media stack 348 level, a new offer can be generated at any point. For example, when 349 using SIP for signaling, if one offer is sent, then cancelled using a 350 SIP CANCEL, another offer can be generated even though no answer was 351 received for the first offer. To support this, the JSEP media layer 352 can provide an offer via the createOffer() method whenever the 353 Javascript application needs one for the signaling. The answerer can 354 send back zero or more provisional answers, and finally end the 355 offer-answer exchange by sending a final answer. The state machine 356 for this is as follows: 358 setRemote(OFFER) setLocal(PRANSWER) 359 /-----\ /-----\ 360 | | | | 361 v | v | 362 +---------------+ | +---------------+ | 363 | |----/ | |----/ 364 | | setLocal(PRANSWER) | | 365 | Remote-Offer |------------------- >| Local-Pranswer| 366 | | | | 367 | | | | 368 +---------------+ +---------------+ 369 ^ | | 370 | | setLocal(ANSWER) | 371 setRemote(OFFER) | | 372 | V setLocal(ANSWER) | 373 +---------------+ | 374 | | | 375 | |<---------------------------+ 376 | Stable | 377 | |<---------------------------+ 378 | | | 379 +---------------+ setRemote(ANSWER) | 380 ^ | | 381 | | setLocal(OFFER) | 382 setRemote(ANSWER) | | 383 | V | 384 +---------------+ +---------------+ 385 | | | | 386 | | setRemote(PRANSWER) | | 387 | Local-Offer |------------------- >|Remote-Pranswer| 388 | | | | 389 | |----\ | |----\ 390 +---------------+ | +---------------+ | 391 ^ | ^ | 392 | | | | 393 \-----/ \-----/ 394 setLocal(OFFER) setRemote(PRANSWER) 396 Figure 2: JSEP State Machine 398 Aside from these state transitions there is no other difference 399 between the handling of provisional ("pranswer") and final ("answer") 400 answers. 402 3.3. Session Description Format 404 In the WebRTC specification, session descriptions are formatted as 405 SDP messages. While this format is not optimal for manipulation from 406 Javascript, it is widely accepted, and frequently updated with new 407 features. Any alternate encoding of session descriptions would have 408 to keep pace with the changes to SDP, at least until the time that 409 this new encoding eclipsed SDP in popularity. As a result, JSEP 410 currently uses SDP as the internal representation for its session 411 descriptions. 413 However, to simplify Javascript processing, and provide for future 414 flexibility, the SDP syntax is encapsulated within a 415 SessionDescription object, which can be constructed from SDP, and be 416 serialized out to SDP. If future specifications agree on a JSON 417 format for session descriptions, we could easily enable this object 418 to generate and consume that JSON. 420 Other methods may be added to SessionDescription in the future to 421 simplify handling of SessionDescriptions from Javascript. In the 422 meantime, Javascript libraries can be used to perform these 423 manipulations. 425 Note that most applications should be able to treat the 426 SessionDescriptions produced and consumed by these various API calls 427 as opaque blobs; that is, the application will not need to read or 428 change them. The W3C WebRTC API specification will provide 429 appropriate APIs to allow the application to control various session 430 parameters, which will provide the necessary information to the 431 browser about what sort of SessionDescription to produce. 433 3.4. ICE 435 When a new ICE candidate is available, the ICE Agent will notify the 436 application via a callback; these candidates will automatically be 437 added to the local session description. When all candidates have 438 been gathered, the callback will also be invoked to signal that the 439 gathering process is complete. 441 3.4.1. ICE Candidate Trickling 443 Candidate trickling is a technique through which a caller may 444 incrementally provide candidates to the callee after the initial 445 offer has been dispatched; the semantics of "Trickle ICE" are defined 446 in [I-D.ietf-mmusic-trickle-ice]. This process allows the callee to 447 begin acting upon the call and setting up the ICE (and perhaps DTLS) 448 connections immediately, without having to wait for the caller to 449 gather all possible candidates. This results in faster media setup 450 in cases where gathering is not performed prior to initiating the 451 call. 453 JSEP supports optional candidate trickling by providing APIs that 454 provide control and feedback on the ICE candidate gathering process. 455 Applications that support candidate trickling can send the initial 456 offer immediately and send individual candidates when they get the 457 notified of a new candidate; applications that do not support this 458 feature can simply wait for the indication that gathering is 459 complete, and then create and send their offer, with all the 460 candidates, at this time. 462 Upon receipt of trickled candidates, the receiving application will 463 supply them to its ICE Agent. This triggers the ICE Agent to start 464 using the new remote candidates for connectivity checks. 466 3.4.1.1. ICE Candidate Format 468 As with session descriptions, the syntax of the IceCandidate object 469 provides some abstraction, but can be easily converted to and from 470 the SDP candidate lines. 472 The candidate lines are the only SDP information that is contained 473 within IceCandidate, as they represent the only information needed 474 that is not present in the initial offer (i.e., for trickle 475 candidates). This information is carried with the same syntax as the 476 "candidate-attribute" field defined for ICE. For example: 478 candidate:1 1 UDP 1694498815 192.0.2.33 10000 typ host 480 The IceCandidate object also contains fields to indicate which m= 481 line it should be associated with. The m line can be identified in 482 one of two ways; either by a m-line index, or a MID. The m-line 483 index is a zero-based index, with index N referring to the N+1th 484 m-line in the SDP sent by the entity which sent the IceCandidate. 485 The MID uses the "media stream identification", as defined in 486 [RFC5888], to identify the m-line. WebRTC implementations creating 487 an ICE Candidate object MUST populate both of these fields. 488 Implementations receiving an ICE Candidate object SHOULD use the MID 489 if they implement that functionality, or the m-line index, if not. 491 3.4.2. ICE Candidate Pool 493 JSEP applications typically inform the browser to begin ICE gathering 494 via the information supplied to setLocalDescription, as this is where 495 the app specifies the number of media streams for which to gather 496 candidates. However, to accelerate cases where the application knows 497 the number of media streams to use ahead of time, it MAY ask the 498 browser to gather a pool of potential ICE candidates to help ensure 499 rapid media setup. When setLocalDescription is eventually called, 500 and the browser goes to gather the needed ICE candidates, it SHOULD 501 start by checking if any candidates are available in the pool. If 502 there are candidates in the pool, they SHOULD be handed to the 503 application immediately via the ICE candidate callback. If the pool 504 becomes depleted, either because a larger-than-expected number of 505 media streams is used, or because the pool has not had enough time to 506 gather candidates, the remaining candidates are gathered as usual. 508 3.5. Interactions With Forking 510 Some call signaling systems allow various types of forking where an 511 SDP Offer may be provided to more than one device. For example, SIP 512 [RFC3261] defines both a "Parallel Search" and "Sequential Search". 513 Although these are primarily signaling level issues that are outside 514 the scope of JSEP, they do have some impact on the configuration of 515 the media plane that is relevant. When forking happens at the 516 signaling layer, the Javascript application responsible for the 517 signaling needs to make the decisions about what media should be sent 518 or received at any point of time, as well as which remote endpoint it 519 should communicate with; JSEP is used to make sure the media engine 520 can make the RTP and media perform as required by the application. 521 The basic operations that the applications can have the media engine 522 do are: 523 o Start exchanging media to a given remote peer, but keep all the 524 resources reserved in the offer. 525 o Start exchanging media with a given remote peer, and free any 526 resources in the offer that are not being used. 528 3.5.1. Sequential Forking 530 Sequential forking involves a call being dispatched to multiple 531 remote callees, where each callee can accept the call, but only one 532 active session ever exists at a time; no mixing of received media is 533 performed. 535 JSEP handles sequential forking well, allowing the application to 536 easily control the policy for selecting the desired remote endpoint. 537 When an answer arrives from one of the callees, the application can 538 choose to apply it either as a provisional answer, leaving open the 539 possibility of using a different answer in the future, or apply it as 540 a final answer, ending the setup flow. 542 In a "first-one-wins" situation, the first answer will be applied as 543 a final answer, and the application will reject any subsequent 544 answers. In SIP parlance, this would be ACK + BYE. 546 In a "last-one-wins" situation, all answers would be applied as 547 provisional answers, and any previous call leg will be terminated. 548 At some point, the application will end the setup process, perhaps 549 with a timer; at this point, the application could reapply the 550 existing remote description as a final answer. 552 3.5.2. Parallel Forking 554 Parallel forking involves a call being dispatched to multiple remote 555 callees, where each callee can accept the call, and multiple 556 simultaneous active signaling sessions can be established as a 557 result. If multiple callees send media at the same time, the 558 possibilities for handling this are described in Section 3.1 of 559 [RFC3960]. Most SIP devices today only support exchanging media with 560 a single device at a time, and do not try to mix multiple early media 561 audio sources, as that could result in a confusing situation. For 562 example, consider having a European ringback tone mixed together with 563 the North American ringback tone - the resulting sound would not be 564 like either tone, and would confuse the user. If the signaling 565 application wishes to only exchange media with one of the remote 566 endpoints at a time, then from a media engine point of view, this is 567 exactly like the sequential forking case. 569 In the parallel forking case where the Javascript application wishes 570 to simultaneously exchange media with multiple peers, the flow is 571 slightly more complex, but the Javascript application can follow the 572 strategy that [RFC3960] describes using UPDATE. The UPDATE approach 573 allows the signaling to set up a separate media flow for each peer 574 that it wishes to exchange media with. In JSEP, this offer used in 575 the UPDATE would be formed by simply creating a new PeerConnection 576 and making sure that the same local media streams have been added 577 into this new PeerConnection. Then the new PeerConnection object 578 would produce a SDP offer that could be used by the signaling to 579 perform the UPDATE strategy discussed in [RFC3960]. 581 As a result of sharing the media streams, the application will end up 582 with N parallel PeerConnection sessions, each with a local and remote 583 description and their own local and remote addresses. The media flow 584 from these sessions can be managed by specifying SDP direction 585 attributes in the descriptions, or the application can choose to play 586 out the media from all sessions mixed together. Of course, if the 587 application wants to only keep a single session, it can simply 588 terminate the sessions that it no longer needs. 590 4. Interface 592 This section details the basic operations that must be present to 593 implement JSEP functionality. The actual API exposed in the W3C API 594 may have somewhat different syntax, but should map easily to these 595 concepts. 597 4.1. Methods 599 4.1.1. Constructor 601 The PeerConnection constructor allows the application to specify 602 global parameters for the media session, such as the STUN/TURN 603 servers and credentials to use when gathering candidates. The size 604 of the ICE candidate pool can also be set, if desired; this indicates 605 the number of ICE components to pre-gather candidates for. If the 606 application does not indicate a candidate pool size, the browser may 607 select any default candidate pool size. 609 In addition, the application can specify its preferred policy 610 regarding use of BUNDLE, the multiplexing mechanism defined in 611 [I-D.ietf-mmusic-sdp-bundle-negotiation]. By specifying a policy 612 from the list below, the application can control how aggressively it 613 will try to BUNDLE media streams together. The set of available 614 policies is as follows: 615 balanced: The application will BUNDLE all media streams of the same 616 type together. That is, if there are multiple audio and multiple 617 video MediaStreamTracks attached to a PeerConnection, all but the 618 first audio and video tracks will be marked as bundle-only, and 619 candidates will only be gathered for N media streams, where N is 620 the number of distinct media types. When talking to a non-BUNDLE- 621 aware endpoint, only the non-bundle-only streams will be 622 negotiated. This policy balances desire to multiplex with the 623 need to ensure basic audio and video still works in legacy cases. 624 Data channels will be in a separate bundle group. 626 max-bundle: The application will BUNDLE all of its media streams, 627 including data channels, on a single transport. All streams other 628 than the first will be marked as bundle-only. This policy aims to 629 minimize candidate gathering and maximize multiplexing, at the 630 cost of less compatibility with legacy endpoints. 632 max-compat: The application will offer BUNDLE, but mark none of its 633 streams as bundle-only. This policy will allow all streams to be 634 received by non-BUNDLE-aware endpoints, but require separate 635 candidates to be gathered for each media stream. 637 4.1.2. createOffer 639 The createOffer method generates a blob of SDP that contains a 640 [RFC3264] offer with the supported configurations for the session, 641 including descriptions of the local MediaStreams attached to this 642 PeerConnection, the codec/RTP/RTCP options supported by this 643 implementation, and any candidates that have been gathered by the ICE 644 Agent. An options parameter may be supplied to provide additional 645 control over the generated offer. This options parameter should 646 allow for the following manipulations to be performed: 648 o To indicate support for a media type even if no MediaStreamTracks 649 of that type have been added to the session (e.g., an audio call 650 that wants to receive video.) 651 o To trigger an ICE restart, for the purpose of reestablishing 652 connectivity. 654 In the initial offer, the generated SDP will contain all desired 655 functionality for the session (functionality that is supported but 656 not desired by default may be omitted); for each SDP line, the 657 generation of the SDP will follow the process defined for generating 658 an initial offer from the document that specifies the given SDP line. 659 The exact handling of initial offer generation is detailed in 660 Section 5.2.1 below. 662 In the event createOffer is called after the session is established, 663 createOffer will generate an offer to modify the current session 664 based on any changes that have been made to the session, e.g. adding 665 or removing MediaStreams, or requesting an ICE restart. For each 666 existing stream, the generation of each SDP line must follow the 667 process defined for generating an updated offer from the RFC that 668 specifies the given SDP line. For each new stream, the generation of 669 the SDP must follow the process of generating an initial offer, as 670 mentioned above. If no changes have been made, or for SDP lines that 671 are unaffected by the requested changes, the offer will only contain 672 the parameters negotiated by the last offer-answer exchange. The 673 exact handling of subsequent offer generation is detailed in 674 Section 5.2.2. below. 676 Session descriptions generated by createOffer must be immediately 677 usable by setLocalDescription; if a system has limited resources 678 (e.g. a finite number of decoders), createOffer should return an 679 offer that reflects the current state of the system, so that 680 setLocalDescription will succeed when it attempts to acquire those 681 resources. Because this method may need to inspect the system state 682 to determine the currently available resources, it may be implemented 683 as an async operation. 685 Calling this method may do things such as generate new ICE 686 credentials, but does not result in candidate gathering, or cause 687 media to start or stop flowing. 689 4.1.3. createAnswer 691 The createAnswer method generates a blob of SDP that contains a 692 [RFC3264] SDP answer with the supported configuration for the session 693 that is compatible with the parameters supplied in the most recent 694 call to setRemoteDescription, which MUST have been called prior to 695 calling createAnswer. Like createOffer, the returned blob contains 696 descriptions of the local MediaStreams attached to this 697 PeerConnection, the codec/RTP/RTCP options negotiated for this 698 session, and any candidates that have been gathered by the ICE Agent. 699 An options parameter may be supplied to provide additional control 700 over the generated answer. 702 As an answer, the generated SDP will contain a specific configuration 703 that specifies how the media plane should be established; for each 704 SDP line, the generation of the SDP must follow the process defined 705 for generating an answer from the document that specifies the given 706 SDP line. The exact handling of answer generation is detailed in 707 Section 5.3. below. 709 Session descriptions generated by createAnswer must be immediately 710 usable by setLocalDescription; like createOffer, the returned 711 description should reflect the current state of the system. Because 712 this method may need to inspect the system state to determine the 713 currently available resources, it may need to be implemented as an 714 async operation. 716 Calling this method may do things such as generate new ICE 717 credentials, but does not trigger candidate gathering or change media 718 state. 720 4.1.4. SessionDescriptionType 722 Session description objects (RTCSessionDescription) may be of type 723 "offer", "pranswer", and "answer". These types provide information 724 as to how the description parameter should be parsed, and how the 725 media state should be changed. 727 "offer" indicates that a description should be parsed as an offer; 728 said description may include many possible media configurations. A 729 description used as an "offer" may be applied anytime the 730 PeerConnection is in a stable state, or as an update to a previously 731 supplied but unanswered "offer". 733 "pranswer" indicates that a description should be parsed as an 734 answer, but not a final answer, and so should not result in the 735 freeing of allocated resources. It may result in the start of media 736 transmission, if the answer does not specify an inactive media 737 direction. A description used as a "pranswer" may be applied as a 738 response to an "offer", or an update to a previously sent "pranswer". 740 "answer" indicates that a description should be parsed as an answer, 741 the offer-answer exchange should be considered complete, and any 742 resources (decoders, candidates) that are no longer needed can be 743 released. A description used as an "answer" may be applied as a 744 response to a "offer", or an update to a previously sent "pranswer". 746 The only difference between a provisional and final answer is that 747 the final answer results in the freeing of any unused resources that 748 were allocated as a result of the offer. As such, the application 749 can use some discretion on whether an answer should be applied as 750 provisional or final, and can change the type of the session 751 description as needed. For example, in a serial forking scenario, an 752 application may receive multiple "final" answers, one from each 753 remote endpoint. The application could choose to accept the initial 754 answers as provisional answers, and only apply an answer as final 755 when it receives one that meets its criteria (e.g. a live user 756 instead of voicemail). 758 "rollback" is a special session description type implying that the 759 state machine should be rolled back to the previous state, as 760 described in Section 4.1.4.2. The contents MUST be empty. 762 4.1.4.1. Use of Provisional Answers 764 Most web applications will not need to create answers using the 765 "pranswer" type. While it is good practice to send an immediate 766 response to an "offer", in order to warm up the session transport and 767 prevent media clipping, the preferred handling for a web application 768 would be to create and send an "inactive" final answer immediately 769 after receiving the offer. Later, when the called user actually 770 accepts the call, the application can create a new "sendrecv" offer 771 to update the previous offer/answer pair and start the media flow. 772 While this could also be done with an inactive "pranswer", followed 773 by a sendrecv "answer", the initial "pranswer" leaves the offer- 774 answer exchange open, which means that neither side can send an 775 updated offer during this time. 777 As an example, consider a typical web application that will set up a 778 data channel, an audio channel, and a video channel. When an 779 endpoint receives an offer with these channels, it could send an 780 answer accepting the data channel for two-way data, and accepting the 781 audio and video tracks as inactive or receive-only. It could then 782 ask the user to accept the call, acquire the local media streams, and 783 send a new offer to the remote side moving the audio and video to be 784 two-way media. By the time the human has accepted the call and 785 triggered the new offer, it is likely that the ICE and DTLS 786 handshaking for all the channels will already have finished. 788 Of course, some applications may not be able to perform this double 789 offer-answer exchange, particularly ones that are attempting to 790 gateway to legacy signaling protocols. In these cases, "pranswer" 791 can still provide the application with a mechanism to warm up the 792 transport. 794 4.1.4.2. Rollback 796 In certain situations it may be desirable to "undo" a change made to 797 setLocalDescription or setRemoteDescription. Consider a case where a 798 call is ongoing, and one side wants to change some of the session 799 parameters; that side generates an updated offer and then calls 800 setLocalDescription. However, the remote side, either before or 801 after setRemoteDescription, decides it does not want to accept the 802 new parameters, and sends a reject message back to the offerer. Now, 803 the offerer, and possibly the answerer as well, need to return to a 804 stable state and the previous local/remote description. To support 805 this, we introduce the concept of "rollback". 807 A rollback discards any proposed changes to the session, returning 808 the state machine to the stable state, and setting the modified local 809 and/or remote description back to their previous values. Any 810 resources or candidates that were allocated by the abandoned local 811 description are discarded; any media that is received will be 812 processed according to the previous local and remote descriptions. 813 Rollback can only be used to cancel proposed changes; there is no 814 support for rolling back from a stable state to a previous stable 815 state. Note that this implies that once the answerer has performed 816 setLocalDescription with his answer, this cannot be rolled back. 818 A rollback is performed by supplying a session description of type 819 "rollback" with empty contents to either setLocalDescription or 820 setRemoteDescription, depending on which was most recently used (i.e. 821 if the new offer was supplied to setLocalDescription, the rollback 822 should be done using setLocalDescription as well). 824 4.1.5. setLocalDescription 826 The setLocalDescription method instructs the PeerConnection to apply 827 the supplied SDP blob as its local configuration. The type field 828 indicates whether the blob should be processed as an offer, 829 provisional answer, or final answer; offers and answers are checked 830 differently, using the various rules that exist for each SDP line. 832 This API changes the local media state; among other things, it sets 833 up local resources for receiving and decoding media. In order to 834 successfully handle scenarios where the application wants to offer to 835 change from one media format to a different, incompatible format, the 836 PeerConnection must be able to simultaneously support use of both the 837 old and new local descriptions (e.g. support codecs that exist in 838 both descriptions) until a final answer is received, at which point 839 the PeerConnection can fully adopt the new local description, or roll 840 back to the old description if the remote side denied the change. 842 This API indirectly controls the candidate gathering process. When a 843 local description is supplied, and the number of transports currently 844 in use does not match the number of transports needed by the local 845 description, the PeerConnection will create transports as needed and 846 begin gathering candidates for them. 848 If setRemoteDescription was previous called with an offer, and 849 setLocalDescription is called with an answer (provisional or final), 850 and the media directions are compatible, and media are available to 851 send, this will result in the starting of media transmission. 853 4.1.6. setRemoteDescription 855 The setRemoteDescription method instructs the PeerConnection to apply 856 the supplied SDP blob as the desired remote configuration. As in 857 setLocalDescription, the type field of the indicates how the blob 858 should be processed. 860 This API changes the local media state; among other things, it sets 861 up local resources for sending and encoding media. 863 If setRemoteDescription was previously called with an offer, and 864 setLocalDescription is called with an answer (provisional or final), 865 and the media directions are compatible, and media are available to 866 send, this will result in the starting of media transmission. 868 4.1.7. localDescription 870 The localDescription method returns a copy of the current local 871 configuration, i.e. what was most recently passed to 872 setLocalDescription, plus any local candidates that have been 873 generated by the ICE Agent. 875 [[OPEN ISSUE: Do we need to expose accessors for both the current 876 and proposed local description? 877 https://github.com/rtcweb-wg/jsep/issues/16]] 879 A null object will be returned if the local description has not yet 880 been established, or if the PeerConnection has been closed. 882 4.1.8. remoteDescription 884 The remoteDescription method returns a copy of the current remote 885 configuration, i.e. what was most recently passed to 886 setRemoteDescription, plus any remote candidates that have been 887 supplied via processIceMessage. 889 [[OPEN ISSUE: Do we need to expose accessors for both the current 890 and proposed remote description? 891 https://github.com/rtcweb-wg/jsep/issues/16]] 893 A null object will be returned if the remote description has not yet 894 been established, or if the PeerConnection has been closed. 896 4.1.9. updateIce 898 The updateIce method allows the configuration of the ICE Agent to be 899 changed during the session, primarily for changing which types of 900 local candidates are provided to the application and used for 901 connectivity checks. A callee may initially configure the ICE Agent 902 to use only relay candidates, to avoid leaking location information, 903 but update this configuration to use all candidates once the call is 904 accepted. 906 Regardless of the configuration, the gathering process collects all 907 available candidates, but excluded candidates will not be surfaced in 908 onicecandidate callback or used for connectivity checks. 910 This call may result in a change to the state of the ICE Agent, and 911 may result in a change to media state if it results in connectivity 912 being established. 914 4.1.10. addIceCandidate 916 The addIceCandidate method provides a remote candidate to the ICE 917 Agent, which, if parsed successfully, will be added to the remote 918 description according to the rules defined for Trickle ICE. 919 Connectivity checks will be sent to the new candidate. 921 This call will result in a change to the state of the ICE Agent, and 922 may result in a change to media state if it results in connectivity 923 being established. 925 5. SDP Interaction Procedures 927 This section describes the specific procedures to be followed when 928 creating and parsing SDP objects. 930 5.1. Requirements Overview 932 JSEP implementations must comply with the specifications listed below 933 that govern the creation and processing of offers and answers. 935 The first set of specifications is the "mandatory-to-implement" set. 936 All implementations must support these behaviors, but may not use all 937 of them if the remote side, which may not be a JSEP endpoint, does 938 not support them. 940 The second set of specifications is the "mandatory-to-use" set. The 941 local JSEP endpoint and any remote endpoint must indicate support for 942 these specifications in their session descriptions. 944 5.1.1. Implementation Requirements 946 This list of mandatory-to-implement specifications is derived from 947 the requirements outlined in [I-D.ietf-rtcweb-rtp-usage]. 948 R-1 [RFC4566] is the base SDP specification and MUST be 949 implemented. 950 R-2 [RFC5764] MUST be supported for signaling the UDP/TLS/RTP/SAVPF 951 RTP profile. 952 R-3 [RFC5245] MUST be implemented for signaling the ICE credentials 953 and candidate lines corresponding to each media stream. The 954 ICE implementation MUST be a Full implementation, not a Lite 955 implementation. 956 R-4 [RFC5763] MUST be implemented to signal DTLS certificate 957 fingerprints. 958 R-5 [RFC4568] MUST NOT be implemented to signal SDES SRTP keying 959 information. 960 R-6 The [RFC5888] grouping framework MUST be implemented for 961 signaling grouping information, and MUST be used to identify m= 962 lines via the a=mid attribute. 963 R-7 [I-D.ietf-mmusic-msid] MUST be supported, in order to signal 964 associations between RTP objects and W3C MediaStreams and 965 MediaStreamTracks in a standard way. 966 R-8 The bundle mechanism in 967 [I-D.ietf-mmusic-sdp-bundle-negotiation] MUST be supported to 968 signal the ability to multiplex RTP streams on a single UDP 969 port, in order to avoid excessive use of port number resources. 970 R-9 The SDP attributes of "sendonly", "recvonly", "inactive", and 971 "sendrecv" from [RFC4566] MUST be implemented to signal 972 information about media direction. 973 R-10 [RFC5576] MUST be implemented to signal RTP SSRC values. 975 R-11 [RFC4585] MUST be implemented to signal RTCP based feedback. 976 R-12 [RFC5761] MUST be implemented to signal multiplexing of RTP and 977 RTCP. 978 R-13 [RFC5506] MUST be implemented to signal reduced-size RTCP 979 messages. 980 R-14 [RFC3556] with bandwidth modifiers MAY be supported for 981 specifying RTCP bandwidth as a fraction of the media bandwidth, 982 RTCP fraction allocated to the senders and setting maximum 983 media bit-rate boundaries. 985 As required by [RFC4566], Section 5.13, JSEP implementations MUST 986 ignore unknown attribute (a=) lines. 988 5.1.2. Usage Requirements 990 All session descriptions handled by JSEP endpoints, both local and 991 remote, MUST indicate support for the following specifications. If 992 any of these are absent, this omission MUST be treated as an error. 993 R-1 Either the UDP/TLS/RTP/SAVP or the UDP/TLS/RTP/SAVPF RTP 994 profile, as specified in [RFC5764], MUST be used. 995 R-2 ICE, as specified in [RFC5245], MUST be used. Note that the 996 remote endpoint may use a Lite implementation; implementations 997 MUST properly handle remote endpoints which do ICE-Lite. 998 R-3 DTLS-SRTP, as specified in [RFC5763], MUST be used. 1000 5.2. Constructing an Offer 1002 When createOffer is called, a new SDP description must be created 1003 that includes the functionality specified in 1004 [I-D.ietf-rtcweb-rtp-usage]. The exact details of this process are 1005 explained below. 1007 5.2.1. Initial Offers 1009 When createOffer is called for the first time, the result is known as 1010 the initial offer. 1012 The first step in generating an initial offer is to generate session- 1013 level attributes, as specified in [RFC4566], Section 5. 1014 Specifically: 1015 o The first SDP line MUST be "v=0", as specified in [RFC4566], 1016 Section 5.1 1017 o The second SDP line MUST be an "o=" line, as specified in 1018 [RFC4566], Section 5.2. The value of the field SHOULD 1019 be "-". The value of the field SHOULD be a 1020 cryptographically random number. To ensure uniqueness, this 1021 number SHOULD be at least 64 bits long. The value of the field SHOULD be zero. The value of the 1023 tuple SHOULD be set to a non- 1024 meaningful address, such as IN IP4 0.0.0.0, to prevent leaking the 1025 local address in this field. As mentioned in [RFC4566], the 1026 entire o= line needs to be unique, but selecting a random number 1027 for is sufficient to accomplish this. 1028 o The third SDP line MUST be a "s=" line, as specified in [RFC4566], 1029 Section 5.3; to match the "o=" line, a single dash SHOULD be used 1030 as the session name, e.g. "s=-". Note that this differs from the 1031 advice in [RFC4566] which proposes a single space, but as both 1032 "o=" and "s=" are meaningless, having the same meaningless value 1033 seems clearer. 1034 o Session Information ("i="), URI ("u="), Email Address ("e="), 1035 Phone Number ("p="), Bandwidth ("b="), Repeat Times ("r="), and 1036 Time Zones ("z=") lines are not useful in this context and SHOULD 1037 NOT be included. 1038 o Encryption Keys ("k=") lines do not provide sufficient security 1039 and MUST NOT be included. 1040 o A "t=" line MUST be added, as specified in [RFC4566], Section 5.9; 1041 both and SHOULD be set to zero, e.g. "t=0 1042 0". 1043 o An "a=msid-semantic:WMS" line MUST be added, as specified in 1044 [I-D.ietf-mmusic-msid], Section 4. 1046 The next step is to generate m= sections, as specified in [RFC4566] 1047 Section 5.14, for each MediaStreamTrack that has been added to the 1048 PeerConnection via the addStream method. (Note that this method 1049 takes a MediaStream, which can contain multiple MediaStreamTracks, 1050 and therefore multiple m= sections can be generated even if addStream 1051 is only called once.) m=sections MUST be sorted first by the order in 1052 which the MediaStreams were added to the PeerConnection, and then by 1053 the alphabetical ordering of the media type for the MediaStreamTrack. 1054 For example, if a MediaStream containing both an audio and a video 1055 MediaStreamTrack is added to a PeerConnection, the resultant m=audio 1056 section will precede the m=video section. If a second MediaStream 1057 containing an audio MediaStreamTrack was added, it would follow the 1058 m=video section. 1060 Each m= section, provided it is not being bundled into another m= 1061 section, MUST generate a unique set of ICE credentials and gather its 1062 own unique set of ICE candidates. Otherwise, it MUST use the same 1063 ICE credentials and candidates as the m= section into which it is 1064 being bundled. Note that this means that for offers, any m= sections 1065 which are not bundle-only MUST have unique ICE credentials and 1066 candidates, since it is possible that the answerer will accept them 1067 without bundling them. 1069 For DTLS, all m= sections MUST use the certificate for the identity 1070 that has been specified for the PeerConnection; as a result, they 1071 MUST all have the same [RFC4572] fingerprint value, or this value 1072 MUST be a session-level attribute. 1074 Each m= section should be generated as specified in [RFC4566], 1075 Section 5.14. For the m= line itself, the following rules MUST be 1076 followed: 1077 o The port value is set to the port of the default ICE candidate for 1078 this m= section; if this m= section is not being bundled into 1079 another m= section, the port value MUST be unique. If no 1080 candidates have yet been gathered, and a 'null' port value is 1081 being used, as indicated in [I-D.ietf-mmusic-trickle-ice], Section 1082 5.1., this port MUST still be unique. 1083 o To properly indicate use of DTLS, the field MUST be set to 1084 "UDP/TLS/RTP/SAVPF", as specified in [RFC5764], Section 8. 1086 Each m= section MUST include the following attribute lines: 1087 o An "a=mid" line, as specified in [RFC5888], Section 4. 1088 o An "a=msid" line, as specified in [I-D.ietf-mmusic-msid], Section 1089 2. 1090 o [OPEN ISSUE: Use of AppID] 1091 o An "a=sendrecv" line, as specified in [RFC3264], Section 5.1. 1092 o For each supported codec, "a=rtpmap" and "a=fmtp" lines, as 1093 specified in [RFC4566], Section 6. For audio, the codecs 1094 specified in [I-D.ietf-rtcweb-audio], Section 3, MUST be be 1095 supported. 1096 o For each primary codec where RTP retransmission should be used, a 1097 corresponding "a=rtpmap" line indicating "rtx" with the clock rate 1098 of the primary codec and an "a=fmtp" line that references the 1099 payload type of the primary codec, as specified in [RFC4588], 1100 Section 8.1. 1101 o For each supported FEC mechanism, a corresponding "a=rtpmap" line 1102 indicating the desired FEC codec. 1103 o "a=ice-ufrag" and "a=ice-passwd" lines, as specified in [RFC5245], 1104 Section 15.4. 1105 o An "a=ice-options" line, with the "trickle" option, as specified 1106 in [I-D.ietf-mmusic-trickle-ice], Section 4. 1107 o For each candidate that has been gathered during the most recent 1108 gathering phase, an "a=candidate" line, as specified in [RFC5245], 1109 Section 4.3., paragraph 3. 1110 o For the current default candidate, a "c=" line, as specified in 1111 [RFC5245], Section 4.3., paragraph 6. If no candidates have been 1112 gathered yet, the default candidate should be set to the 'null' 1113 value defined in [I-D.ietf-mmusic-trickle-ice], Section 5.1. 1114 o An "a=fingerprint" line, as specified in [RFC4572], Section 5; the 1115 algorithm used for the fingerprint MUST match that used in the 1116 certificate signature. 1118 o An "a=setup" line, as specified in [RFC4145], Section 4, and 1119 clarified for use in DTLS-SRTP scenarios in [RFC5763], Section 5. 1120 The role value in the offer MUST be "actpass". 1121 o An "a=rtcp-mux" line, as specified in [RFC5761], Section 5.1.1. 1122 o An "a=rtcp-rsize" line, as specified in [RFC5506], Section 5. 1123 o For each supported RTP header extension, an "a=extmap" line, as 1124 specified in [RFC5285], Section 5. The list of header extensions 1125 that SHOULD/MUST be supported is specified in 1126 [I-D.ietf-rtcweb-rtp-usage], Section 5.2. Any header extensions 1127 that require encryption MUST be specified as indicated in 1128 [RFC6904], Section 4. 1129 o For each supported RTCP feedback mechanism, an "a=rtcp-fb" 1130 mechanism, as specified in [RFC4585], Section 4.2. The list of 1131 RTCP feedback mechanisms that SHOULD/MUST be supported is 1132 specified in [I-D.ietf-rtcweb-rtp-usage], Section 5.1. 1133 o An "a=ssrc" line, as specified in [RFC5576], Section 4.1, 1134 indicating the SSRC to be used for sending media, along with the 1135 mandatory "cname" source attribute, as specified in Section 6.1, 1136 indicating the CNAME for the source. The CNAME must be generated 1137 in accordance with [RFC7022]. [OPEN ISSUE: How are CNAMEs 1138 specified for MSTs? Are they randomly generated for each 1139 MediaStream? If so, can two MediaStreams be synced? See: 1140 https://github.com/rtcweb-wg/jsep/issues/4] 1141 o If RTX is supported for this media type, another "a=ssrc" line 1142 with the RTX SSRC, and an "a=ssrc-group" line, as specified in 1143 [RFC5576], section 4.2, with semantics set to "FID" and including 1144 the primary and RTX SSRCs. 1145 o If FEC is supported for this media type, another "a=ssrc" line 1146 with the FEC SSRC, and an "a=ssrc-group" line, as specified in 1147 [RFC5576], section 4.2, with semantics set to "FEC" and including 1148 the primary and FEC SSRCs. 1149 o [OPEN ISSUE: Handling of a=imageattr] 1150 o If the BUNDLE policy for this PeerConnection is set to "max- 1151 bundle", and this is not the first m= section, or the BUNDLE 1152 policy is set to "default", and this is not the first m= section 1153 for this media type, an "a=bundle-only" line. 1155 Lastly, if a data channel has been created, a m= section MUST be 1156 generated for data. The field MUST be set to "application" 1157 and the field MUST be set to "DTLS/SCTP", as specified in 1158 [I-D.ietf-mmusic-sctp-sdp], Section 3; the "fmt" value MUST be set to 1159 the SCTP port number, as specified in Section 4.1. 1161 Within the data m= section, the "a=mid", "a=ice-ufrag", "a=ice- 1162 passwd", "a=ice-options", "a=candidate", "a=fingerprint", and 1163 "a=setup" lines MUST be included as mentioned above, along with an 1164 "a=sctpmap" line referencing the SCTP port number and specifying the 1165 application protocol indicated in [I-D.ietf-rtcweb-data-protocol]. 1167 [OPEN ISSUE: the -01 of this document is missing this information.] 1169 Once all m= sections have been generated, a session-level "a=group" 1170 attribute MUST be added as specified in [RFC5888]. This attribute 1171 MUST have semantics "BUNDLE", and MUST include the mid identifiers of 1172 each m= section. The effect of this is that the browser offers all 1173 m= sections as one BUNDLE group. However, whether the m= sections 1174 are bundle-only or not depends on the BUNDLE policy. 1176 Attributes which SDP permits to either be at the session level or the 1177 media level SHOULD generally be at the media level even if they are 1178 identical. This promotes readability, especially if one of a set of 1179 initially identical attributes is subsequently changed. 1181 Attributes other than the ones specified above MAY be included, 1182 except for the following attributes which are specifically 1183 incompatible with the requirements of [I-D.ietf-rtcweb-rtp-usage], 1184 and MUST NOT be included: 1185 o "a=crypto" 1186 o "a=key-mgmt" 1187 o "a=ice-lite" 1189 Note that when BUNDLE is used, any additional attributes that are 1190 added MUST follow the advice in [I-D.ietf-mmusic-sdp-mux-attributes] 1191 on how those attributes interact with BUNDLE. 1193 Note that these requirements are in some cases stricter than those of 1194 SDP. Implementations MUST be prepared to accept compliant SDP even 1195 if it would not conform to the requirements for generating SDP in 1196 this specification. 1198 5.2.2. Subsequent Offers 1200 When createOffer is called a second (or later) time, or is called 1201 after a local description has already been installed, the processing 1202 is somewhat different than for an initial offer. 1204 If the initial offer was not applied using setLocalDescription, 1205 meaning the PeerConnection is still in the "stable" state, the steps 1206 for generating an initial offer should be followed, subject to the 1207 following restriction: 1208 o The fields of the "o=" line MUST stay the same except for the 1209 field, which MUST increment if the session 1210 description changes in any way, including the addition of ICE 1211 candidates. 1213 If the initial offer was applied using setLocalDescription, but an 1214 answer from the remote side has not yet been applied, meaning the 1215 PeerConnection is still in the "local-offer" state, an offer is 1216 generated by following the steps in the "stable" state above, along 1217 with these exceptions: 1218 o The "s=" and "t=" lines MUST stay the same. 1219 o Each "a=mid" line MUST stay the same. 1220 o Each "a=ice-ufrag" and "a=ice-pwd" line MUST stay the same unless 1221 the "IceRestart" option (Section 5.2.3 was specified. Note that 1222 it's not clear why you would actually want to do this, since at 1223 this point ICE has not yet started and is thus unlikely to need a 1224 restart. 1225 o For MediaStreamTracks that are still present, the "a=msid", 1226 "a=ssrc", and "a=ssrc-group" lines MUST stay the same. 1227 o If any MediaStreamTracks have been removed, either through the 1228 removeStream method or by removing them from an added MediaStream, 1229 their m= sections MUST be marked as recvonly by changing the value 1230 of the [RFC3264] directional attribute to "a=recvonly". The 1231 "a=msid", "a=ssrc", and "a=ssrc-group" lines MUST be removed from 1232 the associated m= sections. 1233 o If any MediaStreamTracks have been added, and there exist m= 1234 sections of the appropriate media type with no associated 1235 MediaStreamTracks (i.e. as described in the preceding paragraph), 1236 those m= sections MUST be recycled by adding the new 1237 MediaStreamTrack to the m= section. This is done by adding the 1238 necessary "a=msid", "a=ssrc", and "a=ssrc-group" lines to the 1239 recycled m= section, and removing the "a=recvonly" attribute. 1241 If the initial offer was applied using setLocalDescription, and an 1242 answer from the remote side has been applied using 1243 setRemoteDescription, meaning the PeerConnection is in the "remote- 1244 pranswer" or "stable" states, an offer is generated based on the 1245 negotiated session descriptions by following the steps mentioned for 1246 the "local-offer" state above, along with these exceptions: [OPEN 1247 ISSUE: should this be permitted in the remote-pranswer state?] 1248 o If a m= section exists in the current local description, but does 1249 not have an associated local MediaStreamTrack (possibly because 1250 said MediaStreamTrack was removed since the last exchange), a m= 1251 section MUST still be generated in the new offer, as indicated in 1252 [RFC3264], Section 8. The disposition of this section will depend 1253 on the state of the remote MediaStreamTrack associated with this 1254 m= section. If one exists, and it is still in the "live" state, 1255 the new m= section MUST be marked as "a=recvonly", with no 1256 "a=msid" or related attributes present. If no remote 1257 MediaStreamTrack exists, or it is in the "ended" state, the m= 1258 section MUST be marked as rejected, by setting the port to zero, 1259 as indicated in [RFC3264], Section 8.2. 1260 o If any MediaStreamTracks have been added, and there exist recvonly 1261 m= sections of the appropriate media type with no associated 1262 MediaStreamTracks, or rejected m= sections of any media type, 1263 those m= sections MUST be recycled, and a local MediaStreamTrack 1264 associated with these recycled m= sections until all such existing 1265 m= sections have been used. This includes any recvonly or 1266 rejected m= sections created by the preceding paragraph. 1268 In addition, for each non-recycled, non-rejected m= section in the 1269 new offer, the following adjustments are made based on the contents 1270 of the corresponding m= section in the current remote description: 1271 o The m= line and corresponding "a=rtpmap" and "a=fmtp" lines MUST 1272 only include codecs present in the remote description. 1273 o The RTP header extensions MUST only include those that are present 1274 in the remote description. 1275 o The RTCP feedback extensions MUST only include those that are 1276 present in the remote description. 1277 o The "a=rtcp-mux" line MUST only be added if present in the remote 1278 description. 1279 o The "a=rtcp-rsize" line MUST only be added if present in the 1280 remote description. 1282 The "a=group:BUNDLE" attribute MUST include the mid identifiers 1283 specified in the BUNDLE group in the most recent answer, minus any m= 1284 sections that have been marked as rejected, plus any newly added or 1285 re-enabled m= sections. In other words, the BUNDLE attribute must 1286 contain all m= sections that were previously bundled, as long as they 1287 are still alive, as well as any new m= sections. 1289 5.2.3. Options Handling 1291 The createOffer method takes as a parameter an RTCOfferOptions 1292 object. Special processing is performed when generating a SDP 1293 description if the following constraints are present. 1295 5.2.3.1. OfferToReceiveAudio 1297 If the "OfferToReceiveAudio" option is specified, with an integer 1298 value of N, the offer MUST include N non-rejected m= sections with 1299 media type "audio", even if fewer than N audio MediaStreamTracks have 1300 been added to the PeerConnection. This allows the offerer to receive 1301 audio, including multiple independent streams, even when not sending 1302 it; accordingly, the directional attribute on the audio m= sections 1303 without associated MediaStreamTracks MUST be set to recvonly. If 1304 this option is specified in the case where at least N audio 1305 MediaStreamTracks have already been added to the PeerConnection, or N 1306 non-rejected m= sections with media type "audio" would otherwise be 1307 generated, it has no effect. For backwards compatibility, a value of 1308 "true" is interpreted as equivalent to N=1. 1310 5.2.3.2. OfferToReceiveVideo 1312 If the "OfferToReceiveVideo" option is specified, with an integer 1313 value of N, the offer MUST include N non-rejected m= sections with 1314 media type "video", even if fewer than N video MediaStreamTracks have 1315 been added to the PeerConnection. This allows the offerer to receive 1316 video, including multiple independent streams, even when not sending 1317 it; accordingly, the directional attribute on the video m= sections 1318 without associated MediaStreamTracks MUST be set to recvonly. If 1319 this option is specified in the case where at least N video 1320 MediaStreamTracks have already been added to the PeerConnection, or N 1321 non-rejected m= sections with media type "video" would otherwise be 1322 generated, it has no effect. For backwards compatibility, a value of 1323 "true" is interpreted as equivalent to N=1. 1325 5.2.3.3. VoiceActivityDetection 1327 If the "VoiceActivityDetection" option is specified, with a value of 1328 "true", the offer MUST indicate support for silence suppression in 1329 the audio it receives by including comfort noise ("CN") codecs for 1330 each offered audio codec, as specified in [RFC3389], Section 5.1, 1331 except for codecs that have their own internal silence suppression 1332 support. For codecs that have their own internal silence suppression 1333 support, the appropriate fmtp parameters for that codec MUST be 1334 specified to indicate that silence suppression for received audio is 1335 desired. For example, when using the Opus codec, the "usedtx=1" 1336 parameter would be specified in the offer. This option allows the 1337 endpoint to significantly reduce the amount of audio bandwidth it 1338 receives, at the cost of some fidelity, depending on the quality of 1339 the remote VAD algorithm. 1341 5.2.3.4. IceRestart 1343 If the "IceRestart" option is specified, with a value of "true", the 1344 offer MUST indicate an ICE restart by generating new ICE ufrag and 1345 pwd attributes, as specified in RFC5245, Section 9.1.1.1. If this 1346 option is specified on an initial offer, it has no effect (since a 1347 new ICE ufrag and pwd are already generated). This option is useful 1348 for reestablishing connectivity in cases where failures are detected. 1350 5.3. Generating an Answer 1352 When createAnswer is called, a new SDP description must be created 1353 that is compatible with the supplied remote description as well as 1354 the requirements specified in [I-D.ietf-rtcweb-rtp-usage]. The exact 1355 details of this process are explained below. 1357 5.3.1. Initial Answers 1359 When createAnswer is called for the first time after a remote 1360 description has been provided, the result is known as the initial 1361 answer. If no remote description has been installed, an answer 1362 cannot be generated, and an error MUST be returned. 1364 Note that the remote description SDP may not have been created by a 1365 JSEP endpoint and may not conform to all the requirements listed in 1366 Section 5.2. For many cases, this is not a problem. However, if any 1367 mandatory SDP attributes are missing, or functionality listed as 1368 mandatory-to-use above is not present, this MUST be treated as an 1369 error, and MUST cause the affected m= sections to be marked as 1370 rejected. 1372 The first step in generating an initial answer is to generate 1373 session-level attributes. The process here is identical to that 1374 indicated in the Initial Offers section above. 1376 The next step is to generate m= sections for each m= section that is 1377 present in the remote offer, as specified in [RFC3264], Section 6. 1378 For the purposes of this discussion, any session-level attributes in 1379 the offer that are also valid as media-level attributes SHALL be 1380 considered to be present in each m= section. 1382 The next step is to go through each offered m= section. If there is 1383 a local MediaStreamTrack of the same type which has been added to the 1384 PeerConnection via addStream and not yet associated with a m= 1385 section, and the specific m= section is either sendrecv or recvonly, 1386 the MediaStreamTrack will be associated with the m= section at this 1387 time. MediaStreamTracks are assigned to m= sections using the 1388 canonical order described in Section 5.2.1. If there are more m= 1389 sections of a certain type than MediaStreamTracks, some m= sections 1390 will not have an associated MediaStreamTrack. If there are more 1391 MediaStreamTracks of a certain type than compatible m= sections, only 1392 the first N MediaStreamTracks will be able to be associated in the 1393 constructed answer. The remainder will need to be associated in a 1394 subsequent offer. 1396 For each offered m= section, if the associated remote 1397 MediaStreamTrack has been stopped, and is therefore in state "ended", 1398 and no local MediaStreamTrack has been associated, the corresponding 1399 m= section in the answer MUST be marked as rejected by setting the 1400 port in the m= line to zero, as indicated in [RFC3264], Section 6., 1401 and further processing for this m= section can be skipped. 1403 Provided that is not the case, each m= section in the answer should 1404 then be generated as specified in [RFC3264], Section 6.1. Because 1405 use of DTLS is mandatory, the field MUST be set to "UDP/TLS/ 1406 RTP/SAVPF". If the offer supports BUNDLE, all m= sections to be 1407 BUNDLEd must use the same ICE credentials and candidates; all m= 1408 sections not being BUNDLEd must use unique ICE credentials and 1409 candidates. Each m= section MUST include the following: 1410 o If present in the offer, an "a=mid" line, as specified in 1411 [RFC5888], Section 9.1. The "mid" value MUST match that specified 1412 in the offer. 1413 o If a local MediaStreamTrack has been associated, an "a=msid" line, 1414 as specified in [I-D.ietf-mmusic-msid], Section 2. 1415 o [OPEN ISSUE: Use of AppID] 1416 o Depending on the directionality of the offer, the disposition of 1417 any associated remote MediaStreamTrack, and the presence of an 1418 associated local MediaStreamTrack, the appropriate directionality 1419 attribute, as specified in [RFC3264], Section 6.1. If the offer 1420 was sendrecv, and the remote MediaStreamTrack is still "live", and 1421 there is a local MediaStreamTrack that has been associated, the 1422 directionality MUST be set as sendrecv. If the offer was 1423 sendonly, and the remote MediaStreamTrack is still "live", the 1424 directionality MUST be set as recvonly. If the offer was 1425 recvonly, and a local MediaStreamTrack has been associated, the 1426 directionality MUST be set as sendonly. If the offer was 1427 inactive, the directionality MUST be set as inactive. 1428 o For each supported codec that is present in the offer, "a=rtpmap" 1429 and "a=fmtp" lines, as specified in [RFC4566], Section 6, and 1430 [RFC3264], Section 6.1. For audio, the codecs specified in 1431 [I-D.ietf-rtcweb-audio], Section 3, MUST be supported. Note that 1432 for simplicity, the answerer MAY use different payload types for 1433 codecs than the offerer, as it is not prohibited by Section 6.1. 1434 o If "rtx" is present in the offer, for each primary codec where RTP 1435 retransmission should be used, a corresponding "a=rtpmap" line 1436 indicating "rtx" with the clock rate of the primary codec and an 1437 "a=fmtp" line that references the payload type of the primary 1438 codec, as specified in [RFC4588], Section 8.1. 1439 o For each supported FEC mechanism that is present in the offer, a 1440 corresponding "a=rtpmap" line indicating the desired FEC codec. 1441 o "a=ice-ufrag" and "a=ice-passwd" lines, as specified in [RFC5245], 1442 Section 15.4. 1443 o If the "trickle" ICE option is present in the offer, an "a=ice- 1444 options" line, with the "trickle" option, as specified in 1445 [I-D.ietf-mmusic-trickle-ice], Section 4. 1446 o For each candidate that has been gathered during the most recent 1447 gathering phase, an "a=candidate" line, as specified in [RFC5245], 1448 Section 4.3., paragraph 3. 1449 o For the current default candidate, a "c=" line, as specified in 1450 [RFC5245], Section 4.3., paragraph 6. If no candidates have been 1451 gathered yet, the default candidate should be set to the 'null' 1452 value defined in [I-D.ietf-mmusic-trickle-ice], Section 5.1. 1454 o An "a=fingerprint" line, as specified in [RFC4572], Section 5; the 1455 algorithm used for the fingerprint MUST match that used in the 1456 certificate signature. 1457 o An "a=setup" line, as specified in [RFC4145], Section 4, and 1458 clarified for use in DTLS-SRTP scenarios in [RFC5763], Section 5. 1459 The role value in the answer MUST be "active" or "passive"; the 1460 "active" role is RECOMMENDED. 1461 o If present in the offer, an "a=rtcp-mux" line, as specified in 1462 [RFC5761], Section 5.1.1. 1463 o If present in the offer, an "a=rtcp-rsize" line, as specified in 1464 [RFC5506], Section 5. 1465 o For each supported RTP header extension that is present in the 1466 offer, an "a=extmap" line, as specified in [RFC5285], Section 5. 1467 The list of header extensions that SHOULD/MUST be supported is 1468 specified in [I-D.ietf-rtcweb-rtp-usage], Section 5.2. Any header 1469 extensions that require encryption MUST be specified as indicated 1470 in [RFC6904], Section 4. 1471 o For each supported RTCP feedback mechanism that is present in the 1472 offer, an "a=rtcp-fb" mechanism, as specified in [RFC4585], 1473 Section 4.2. The list of RTCP feedback mechanisms that SHOULD/ 1474 MUST be supported is specified in [I-D.ietf-rtcweb-rtp-usage], 1475 Section 5.1. 1476 o If a local MediaStreamTrack has been associated, an "a=ssrc" line, 1477 as specified in [RFC5576], Section 4.1, indicating the SSRC to be 1478 used for sending media. 1479 o If a local MediaStreamTrack has been associated, and RTX has been 1480 negotiated for this m= section, another "a=ssrc" line with the RTX 1481 SSRC, and an "a=ssrc-group" line, as specified in [RFC5576], 1482 section 4.2, with semantics set to "FID" and including the primary 1483 and RTX SSRCs. 1484 o If a local MediaStreamTrack has been associated, and FEC has been 1485 negotiated for this m= section, another "a=ssrc" line with the FEC 1486 SSRC, and an "a=ssrc-group" line, as specified in [RFC5576], 1487 section 4.2, with semantics set to "FEC" and including the primary 1488 and FEC SSRCs. 1489 o [OPEN ISSUE: Handling of a=imageattr] 1491 If a data channel m= section has been offered, a m= section MUST also 1492 be generated for data. The field MUST be set to 1493 "application" and the field MUST be set to "DTLS/SCTP", as 1494 specified in [I-D.ietf-mmusic-sctp-sdp], Section 3; the "fmt" value 1495 MUST be set to the SCTP port number, as specified in Section 4.1. 1497 Within the data m= section, the "a=mid", "a=ice-ufrag", "a=ice- 1498 passwd", "a=ice-options", "a=candidate", "a=fingerprint", and 1499 "a=setup" lines MUST be included as mentioned above, along with an 1500 "a=sctpmap" line referencing the SCTP port number and specifying the 1501 application protocol indicated in [I-D.ietf-rtcweb-data-protocol]. 1503 [OPEN ISSUE: the -01 of this document is missing this information.] 1505 If "a=group" attributes with semantics of "BUNDLE" are offered, 1506 corresponding session-level "a=group" attributes MUST be added as 1507 specified in [RFC5888]. These attributes MUST have semantics 1508 "BUNDLE", and MUST include the all mid identifiers from the offered 1509 BUNDLE groups that have not been rejected. Note that regardless of 1510 the presence of "a=bundle-only" in the offer, no m= sections in the 1511 answer should have an "a=bundle-only" line. 1513 Attributes that are common between all m= sections MAY be moved to 1514 session-level, if explicitly defined to be valid at session-level. 1516 The attributes prohibited in the creation of offers are also 1517 prohibited in the creation of answers. 1519 5.3.2. Subsequent Answers 1521 5.3.3. Options Handling 1523 5.4. Parsing an Offer 1525 5.5. Parsing an Answer 1527 5.6. Applying a Local Description 1529 5.7. Applying a Remote Description 1531 6. Configurable SDP Parameters 1533 It is possible to change elements in the SDP returned from 1534 createOffer before passing it to setLocalDescription. When an 1535 implementation receives modified SDP it MUST either: 1537 o Accept the changes and adjust its behavior to match the SDP. 1538 o Reject the changes and return an error via the error callback. 1540 Changes MUST NOT be silently ignored. 1542 The following elements of the SDP media description MUST NOT be 1543 changed between the createOffer and the setLocalDescription, since 1544 they reflect transport attributes that are solely under browser 1545 control, and the browser MUST NOT honor an attempt to change them: 1547 o The number, type and port number of m-lines. 1549 o The generated ICE credentials (a=ice-ufrag and a=ice-pwd). 1550 o The set of ICE candidates and their parameters (a=candidate). 1552 The following modifications, if done by the browser to a description 1553 between createOffer/createAnswer and the setLocalDescription, MUST be 1554 honored by the browser: 1556 o Remove or reorder codecs (m=) 1558 The following parameters may be controlled by constraints passed into 1559 createOffer/createAnswer. As an open issue, these changes may also 1560 be be performed by manipulating the SDP returned from createOffer/ 1561 createAnswer, as indicated above, as long as the capabilities of the 1562 endpoint are not exceeded (e.g. asking for a resolution greater than 1563 what the endpoint can encode): 1565 o [[OPEN ISSUE: This is a placeholder for other modifications, 1566 which we may continue adding as use cases appear.]] 1568 Implementations MAY choose to either honor or reject any elements not 1569 listed in the above two categories, but must do so explicitly as 1570 described at the beginning of this section. Note that future 1571 standards may add new SDP elements to the list of elements which must 1572 be accepted or rejected, but due to version skew, applications must 1573 be prepared for implementations to accept changes which must be 1574 rejected and vice versa. 1576 The application can also modify the SDP to reduce the capabilities in 1577 the offer it sends to the far side or the offer that it installs from 1578 the far side in any way the application sees fit, as long as it is a 1579 valid SDP offer and specifies a subset of what was in the original 1580 offer. This is safe because the answer is not permitted to expand 1581 capabilities and therefore will just respond to what is actually in 1582 the offer. 1584 As always, the application is solely responsible for what it sends to 1585 the other party, and all incoming SDP will be processed by the 1586 browser to the extent of its capabilities. It is an error to assume 1587 that all SDP is well-formed; however, one should be able to assume 1588 that any implementation of this specification will be able to 1589 process, as a remote offer or answer, unmodified SDP coming from any 1590 other implementation of this specification. 1592 7. Security Considerations 1594 The IETF has published separate documents 1595 [I-D.ietf-rtcweb-security-arch] [I-D.ietf-rtcweb-security] describing 1596 the security architecture for WebRTC as a whole. The remainder of 1597 this section describes security considerations for this document. 1599 While formally the JSEP interface is an API, it is better to think of 1600 it is an Internet protocol, with the JS being untrustworthy from the 1601 perspective of the browser. Thus, the threat model of [RFC3552] 1602 applies. In particular, JS can call the API in any order and with 1603 any inputs, including malicious ones. This is particularly relevant 1604 when we consider the SDP which is passed to setLocalDescription(). 1605 While correct API usage requires that the application pass in SDP 1606 which was derived from createOffer() or createAnswer() (perhaps 1607 suitably modified as described in Section 6, there is no guarantee 1608 that applications do so. The browser MUST be prepared for the JS to 1609 pass in bogus data instead. 1611 Conversely, the application programmer MUST recognize that the JS 1612 does not have complete control of browser behavior. One case that 1613 bears particular mention is that editing ICE candidates out of the 1614 SDP or suppressing trickled candidates does not have the expected 1615 behavior: implementations will still perform checks from those 1616 candidates even if they are not sent to the other side. Thus, for 1617 instance, it is not possible to prevent the remote peer from learning 1618 your public IP address by removing server reflexive candidates. 1619 Applications which wish to conceal their public IP address should 1620 instead configure the ICE agent to use only relay candidates. 1622 8. IANA Considerations 1624 This document requires no actions from IANA. 1626 9. Acknowledgements 1628 Significant text incorporated in the draft as well and review was 1629 provided by Harald Alvestrand and Suhas Nandakumar. Dan Burnett, 1630 Neil Stratford, Eric Rescorla, Anant Narayanan, Andrew Hutton, 1631 Richard Ejzak, Adam Bergkvist and Matthew Kaufman all provided 1632 valuable feedback on this proposal. 1634 10. References 1636 10.1. Normative References 1638 [I-D.ietf-mmusic-msid] 1639 Alvestrand, H., "Cross Session Stream Identification in 1640 the Session Description Protocol", 1641 draft-ietf-mmusic-msid-01 (work in progress), August 2013. 1643 [I-D.ietf-mmusic-sctp-sdp] 1644 Loreto, S. and G. Camarillo, "Stream Control Transmission 1645 Protocol (SCTP)-Based Media Transport in the Session 1646 Description Protocol (SDP)", draft-ietf-mmusic-sctp-sdp-04 1647 (work in progress), June 2013. 1649 [I-D.ietf-mmusic-sdp-bundle-negotiation] 1650 Holmberg, C., Alvestrand, H., and C. Jennings, 1651 "Multiplexing Negotiation Using Session Description 1652 Protocol (SDP) Port Numbers", 1653 draft-ietf-mmusic-sdp-bundle-negotiation-04 (work in 1654 progress), June 2013. 1656 [I-D.ietf-mmusic-sdp-mux-attributes] 1657 Nandakumar, S., "A Framework for SDP Attributes when 1658 Multiplexing", draft-ietf-mmusic-sdp-mux-attributes-01 1659 (work in progress), February 2014. 1661 [I-D.ietf-rtcweb-audio] 1662 Valin, J. and C. Bran, "WebRTC Audio Codec and Processing 1663 Requirements", draft-ietf-rtcweb-audio-02 (work in 1664 progress), August 2013. 1666 [I-D.ietf-rtcweb-data-protocol] 1667 Jesup, R., Loreto, S., and M. Tuexen, "WebRTC Data Channel 1668 Protocol", draft-ietf-rtcweb-data-protocol-04 (work in 1669 progress), February 2013. 1671 [I-D.ietf-rtcweb-rtp-usage] 1672 Perkins, C., Westerlund, M., and J. Ott, "Web Real-Time 1673 Communication (WebRTC): Media Transport and Use of RTP", 1674 draft-ietf-rtcweb-rtp-usage-09 (work in progress), 1675 September 2013. 1677 [I-D.ietf-rtcweb-security] 1678 Rescorla, E., "Security Considerations for WebRTC", 1679 draft-ietf-rtcweb-security-06 (work in progress), 1680 January 2014. 1682 [I-D.ietf-rtcweb-security-arch] 1683 Rescorla, E., "WebRTC Security Architecture", 1684 draft-ietf-rtcweb-security-arch-09 (work in progress), 1685 February 2014. 1687 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1688 Requirement Levels", BCP 14, RFC 2119, March 1997. 1690 [RFC3261] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, 1691 A., Peterson, J., Sparks, R., Handley, M., and E. 1692 Schooler, "SIP: Session Initiation Protocol", RFC 3261, 1693 June 2002. 1695 [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model 1696 with Session Description Protocol (SDP)", RFC 3264, 1697 June 2002. 1699 [RFC3552] Rescorla, E. and B. Korver, "Guidelines for Writing RFC 1700 Text on Security Considerations", BCP 72, RFC 3552, 1701 July 2003. 1703 [RFC4145] Yon, D. and G. Camarillo, "TCP-Based Media Transport in 1704 the Session Description Protocol (SDP)", RFC 4145, 1705 September 2005. 1707 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 1708 Description Protocol", RFC 4566, July 2006. 1710 [RFC4572] Lennox, J., "Connection-Oriented Media Transport over the 1711 Transport Layer Security (TLS) Protocol in the Session 1712 Description Protocol (SDP)", RFC 4572, July 2006. 1714 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 1715 "Extended RTP Profile for Real-time Transport Control 1716 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, 1717 July 2006. 1719 [RFC5124] Ott, J. and E. Carrara, "Extended Secure RTP Profile for 1720 Real-time Transport Control Protocol (RTCP)-Based Feedback 1721 (RTP/SAVPF)", RFC 5124, February 2008. 1723 [RFC5245] Rosenberg, J., "Interactive Connectivity Establishment 1724 (ICE): A Protocol for Network Address Translator (NAT) 1725 Traversal for Offer/Answer Protocols", RFC 5245, 1726 April 2010. 1728 [RFC5285] Singer, D. and H. Desineni, "A General Mechanism for RTP 1729 Header Extensions", RFC 5285, July 2008. 1731 [RFC5761] Perkins, C. and M. Westerlund, "Multiplexing RTP Data and 1732 Control Packets on a Single Port", RFC 5761, April 2010. 1734 [RFC5888] Camarillo, G. and H. Schulzrinne, "The Session Description 1735 Protocol (SDP) Grouping Framework", RFC 5888, June 2010. 1737 [RFC6904] Lennox, J., "Encryption of Header Extensions in the Secure 1738 Real-time Transport Protocol (SRTP)", RFC 6904, 1739 April 2013. 1741 [RFC7022] Begen, A., Perkins, C., Wing, D., and E. Rescorla, 1742 "Guidelines for Choosing RTP Control Protocol (RTCP) 1743 Canonical Names (CNAMEs)", RFC 7022, September 2013. 1745 10.2. Informative References 1747 [I-D.ietf-mmusic-trickle-ice] 1748 Ivov, E., Rescorla, E., and J. Uberti, "Trickle ICE: 1749 Incremental Provisioning of Candidates for the Interactive 1750 Connectivity Establishment (ICE) Protocol", 1751 draft-ietf-mmusic-trickle-ice-00 (work in progress), 1752 March 2013. 1754 [I-D.nandakumar-rtcweb-sdp] 1755 Nandakumar, S. and C. Jennings, "SDP for the WebRTC", 1756 draft-nandakumar-rtcweb-sdp-02 (work in progress), 1757 July 2013. 1759 [RFC3389] Zopf, R., "Real-time Transport Protocol (RTP) Payload for 1760 Comfort Noise (CN)", RFC 3389, September 2002. 1762 [RFC3556] Casner, S., "Session Description Protocol (SDP) Bandwidth 1763 Modifiers for RTP Control Protocol (RTCP) Bandwidth", 1764 RFC 3556, July 2003. 1766 [RFC3960] Camarillo, G. and H. Schulzrinne, "Early Media and Ringing 1767 Tone Generation in the Session Initiation Protocol (SIP)", 1768 RFC 3960, December 2004. 1770 [RFC4568] Andreasen, F., Baugher, M., and D. Wing, "Session 1771 Description Protocol (SDP) Security Descriptions for Media 1772 Streams", RFC 4568, July 2006. 1774 [RFC4588] Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R. 1775 Hakenberg, "RTP Retransmission Payload Format", RFC 4588, 1776 July 2006. 1778 [RFC5506] Johansson, I. and M. Westerlund, "Support for Reduced-Size 1779 Real-Time Transport Control Protocol (RTCP): Opportunities 1780 and Consequences", RFC 5506, April 2009. 1782 [RFC5576] Lennox, J., Ott, J., and T. Schierl, "Source-Specific 1783 Media Attributes in the Session Description Protocol 1784 (SDP)", RFC 5576, June 2009. 1786 [RFC5763] Fischl, J., Tschofenig, H., and E. Rescorla, "Framework 1787 for Establishing a Secure Real-time Transport Protocol 1788 (SRTP) Security Context Using Datagram Transport Layer 1789 Security (DTLS)", RFC 5763, May 2010. 1791 [RFC5764] McGrew, D. and E. Rescorla, "Datagram Transport Layer 1792 Security (DTLS) Extension to Establish Keys for the Secure 1793 Real-time Transport Protocol (SRTP)", RFC 5764, May 2010. 1795 [W3C.WD-webrtc-20140617] 1796 Bergkvist, A., Burnett, D., Narayanan, A., and C. 1797 Jennings, "WebRTC 1.0: Real-time Communication Between 1798 Browsers", World Wide Web Consortium WD WD-webrtc- 1799 20140617, June 2014, 1800 . 1802 Appendix A. JSEP Implementation Examples 1804 A.1. Example API Flows 1806 Below are several sample flows for the new PeerConnection and library 1807 APIs, demonstrating when the various APIs are called in different 1808 situations and with various transport protocols. For clarity and 1809 simplicity, the createOffer/createAnswer calls are assumed to be 1810 synchronous in these examples, whereas the actual APIs are async. 1812 A.1.1. Call using ROAP 1814 This example demonstrates a ROAP call, without the use of trickle 1815 candidates. 1817 // Call is initiated toward Answerer 1818 OffererJS->OffererUA: pc = new PeerConnection(); 1819 OffererJS->OffererUA: pc.addStream(localStream, null); 1820 OffererUA->OffererJS: iceCallback(candidate); 1821 OffererJS->OffererUA: offer = pc.createOffer(null); 1822 OffererJS->OffererUA: pc.setLocalDescription("offer", offer); 1823 OffererJS->AnswererJS: {"type":"OFFER", "sdp":offer } 1825 // OFFER arrives at Answerer 1826 AnswererJS->AnswererUA: pc = new PeerConnection(); 1827 AnswererJS->AnswererUA: pc.setRemoteDescription("offer", msg.sdp); 1828 AnswererUA->AnswererJS: onaddstream(remoteStream); 1829 AnswererUA->OffererUA: iceCallback(candidate); 1831 // Answerer accepts call 1832 AnswererJS->AnswererUA: pc.addStream(localStream, null); 1833 AnswererJS->AnswererUA: answer = pc.createAnswer(msg.sdp, null); 1834 AnswererJS->AnswererUA: pc.setLocalDescription("answer", answer); 1835 AnswererJS->OffererJS: {"type":"ANSWER","sdp":answer } 1837 // ANSWER arrives at Offerer 1838 OffererJS->OffererUA: pc.setRemoteDescription("answer", answer); 1839 OffererUA->OffererJS: onaddstream(remoteStream); 1841 // ICE Completes (at Answerer) 1842 AnswererUA->OffererUA: Media 1844 // ICE Completes (at Offerer) 1845 OffererJS->AnswererJS: {"type":"OK" } 1846 OffererUA->AnswererUA: Media 1848 A.1.2. Call using XMPP 1850 This example demonstrates an XMPP call, making use of trickle 1851 candidates. 1853 // Call is initiated toward Answerer 1854 OffererJS->OffererUA: pc = new PeerConnection(); 1855 OffererJS->OffererUA: pc.addStream(localStream, null); 1856 OffererJS->OffererUA: offer = pc.createOffer(null); 1857 OffererJS->OffererUA: pc.setLocalDescription("offer", offer); 1858 OffererJS: xmpp = createSessionInitiate(offer); 1859 OffererJS->AnswererJS: 1861 OffererJS->OffererUA: pc.startIce(); 1862 OffererUA->OffererJS: onicecandidate(cand); 1863 OffererJS: createTransportInfo(cand); 1864 OffererJS->AnswererJS: 1866 // session-initiate arrives at Answerer 1867 AnswererJS->AnswererUA: pc = new PeerConnection(); 1868 AnswererJS: offer = parseSessionInitiate(xmpp); 1869 AnswererJS->AnswererUA: pc.setRemoteDescription("offer", offer); 1870 AnswererUA->AnswererJS: onaddstream(remoteStream); 1872 // transport-infos arrive at Answerer 1873 AnswererJS->AnswererUA: candidate = parseTransportInfo(xmpp); 1874 AnswererJS->AnswererUA: pc.addIceCandidate(candidate); 1875 AnswererUA->AnswererJS: onicecandidate(cand) 1876 AnswererJS: createTransportInfo(cand); 1877 AnswererJS->OffererJS: 1879 // transport-infos arrive at Offerer 1880 OffererJS->OffererUA: candidates = parseTransportInfo(xmpp); 1881 OffererJS->OffererUA: pc.addIceCandidate(candidates); 1883 // Answerer accepts call 1884 AnswererJS->AnswererUA: pc.addStream(localStream, null); 1885 AnswererJS->AnswererUA: answer = pc.createAnswer(offer, null); 1886 AnswererJS: xmpp = createSessionAccept(answer); 1887 AnswererJS->AnswererUA: pc.setLocalDescription("answer", answer); 1888 AnswererJS->OffererJS: 1890 // session-accept arrives at Offerer 1891 OffererJS: answer = parseSessionAccept(xmpp); 1892 OffererJS->OffererUA: pc.setRemoteDescription("answer", answer); 1893 OffererUA->OffererJS: onaddstream(remoteStream); 1895 // ICE Completes (at Answerer) 1896 AnswererUA->OffererUA: Media 1898 // ICE Completes (at Offerer) 1899 OffererUA->AnswererUA: Media 1901 A.1.3. Adding video to a call, using XMPP 1903 This example demonstrates an XMPP call, where the XMPP content-add 1904 mechanism is used to add video media to an existing session. For 1905 simplicity, candidate exchange is not shown. 1907 Note that the offerer for the change to the session may be different 1908 than the original call offerer. 1910 // Offerer adds video stream 1911 OffererJS->OffererUA: pc.addStream(videoStream) 1912 OffererJS->OffererUA: offer = pc.createOffer(null); 1913 OffererJS: xmpp = createContentAdd(offer); 1914 OffererJS->OffererUA: pc.setLocalDescription("offer", offer); 1915 OffererJS->AnswererJS: 1917 // content-add arrives at Answerer 1918 AnswererJS: offer = parseContentAdd(xmpp); 1919 AnswererJS->AnswererUA: pc.setRemoteDescription("offer", offer); 1920 AnswererJS->AnswererUA: answer = pc.createAnswer(offer, null); 1921 AnswererJS->AnswererUA: pc.setLocalDescription("answer", answer); 1922 AnswererJS: xmpp = createContentAccept(answer); 1923 AnswererJS->OffererJS: 1925 // content-accept arrives at Offerer 1926 OffererJS: answer = parseContentAccept(xmpp); 1927 OffererJS->OffererUA: pc.setRemoteDescription("answer", answer); 1929 A.1.4. Simultaneous add of video streams, using XMPP 1931 This example demonstrates an XMPP call, where new video sources are 1932 added at the same time to a call that already has video; since adding 1933 these sources only affects one side of the call, there is no 1934 conflict. The XMPP description-info mechanism is used to indicate 1935 the new sources to the remote side. 1937 // Offerer and "Answerer" add video streams at the same time 1938 OffererJS->OffererUA: pc.addStream(offererVideoStream2) 1939 OffererJS->OffererUA: offer = pc.createOffer(null); 1940 OffererJS: xmpp = createDescriptionInfo(offer); 1941 OffererJS->OffererUA: pc.setLocalDescription("offer", offer); 1942 OffererJS->AnswererJS: 1944 AnswererJS->AnswererUA: pc.addStream(answererVideoStream2) 1945 AnswererJS->AnswererUA: offer = pc.createOffer(null); 1946 AnswererJS: xmpp = createDescriptionInfo(offer); 1947 AnswererJS->AnswererUA: pc.setLocalDescription("offer", offer); 1948 AnswererJS->OffererJS: 1950 // description-info arrives at "Answerer", and is acked 1951 AnswererJS: offer = parseDescriptionInfo(xmpp); 1952 AnswererJS->OffererJS: // ack 1954 // description-info arrives at Offerer, and is acked 1955 OffererJS: offer = parseDescriptionInfo(xmpp); 1956 OffererJS->AnswererJS: // ack 1958 // ack arrives at Offerer; remote offer is used as an answer 1959 OffererJS->OffererUA: pc.setRemoteDescription("answer", offer); 1961 // ack arrives at "Answerer"; remote offer is used as an answer 1962 AnswererJS->AnswererUA: pc.setRemoteDescription("answer", offer); 1964 A.1.5. Call using SIP 1966 This example demonstrates a simple SIP call (e.g. where the client 1967 talks to a SIP proxy over WebSockets). 1969 // Call is initiated toward Answerer 1970 OffererJS->OffererUA: pc = new PeerConnection(); 1971 OffererJS->OffererUA: pc.addStream(localStream, null); 1972 OffererUA->OffererJS: onicecandidate(candidate); 1973 OffererJS->OffererUA: offer = pc.createOffer(null); 1974 OffererJS->OffererUA: pc.setLocalDescription("offer", offer); 1975 OffererJS: sip = createInvite(offer); 1976 OffererJS->AnswererJS: SIP INVITE w/ SDP 1978 // INVITE arrives at Answerer 1979 AnswererJS->AnswererUA: pc = new PeerConnection(); 1980 AnswererJS: offer = parseInvite(sip); 1981 AnswererJS->AnswererUA: pc.setRemoteDescription("offer", offer); 1982 AnswererUA->AnswererJS: onaddstream(remoteStream); 1983 AnswererUA->OffererUA: onicecandidate(candidate); 1985 // Answerer accepts call 1986 AnswererJS->AnswererUA: pc.addStream(localStream, null); 1987 AnswererJS->AnswererUA: answer = pc.createAnswer(offer, null); 1988 AnswererJS: sip = createResponse(200, answer); 1989 AnswererJS->AnswererUA: pc.setLocalDescription("answer", answer); 1990 AnswererJS->OffererJS: 200 OK w/ SDP 1992 // 200 OK arrives at Offerer 1993 OffererJS: answer = parseResponse(sip); 1994 OffererJS->OffererUA: pc.setRemoteDescription("answer", answer); 1995 OffererUA->OffererJS: onaddstream(remoteStream); 1996 OffererJS->AnswererJS: ACK 1998 // ICE Completes (at Answerer) 1999 AnswererUA->OffererUA: Media 2001 // ICE Completes (at Offerer) 2002 OffererUA->AnswererUA: Media 2004 A.1.6. Handling early media (e.g. 1-800-GO FEDEX), using SIP 2006 This example demonstrates how early media could be handled; for 2007 simplicity, only the offerer side of the call is shown. 2009 // Call is initiated toward Answerer 2010 OffererJS->OffererUA: pc = new PeerConnection(); 2011 OffererJS->OffererUA: pc.addStream(localStream, null); 2012 OffererUA->OffererJS: onicecandidate(candidate); 2013 OffererJS->OffererUA: offer = pc.createOffer(null); 2014 OffererJS->OffererUA: pc.setLocalDescription("offer", offer); 2015 OffererJS: sip = createInvite(offer); 2016 OffererJS->AnswererJS: SIP INVITE w/ SDP 2018 // 180 Ringing is received by offerer, w/ SDP 2019 OffererJS: answer = parseResponse(sip); 2020 OffererJS->OffererUA: pc.setRemoteDescription("pranswer", answer); 2021 OffererUA->OffererJS: onaddstream(remoteStream); 2023 // ICE Completes (at Offerer) 2024 OffererUA->AnswererUA: Media 2026 // 200 OK arrives at Offerer 2027 OffererJS: answer = parseResponse(sip); 2028 OffererJS->OffererUA: pc.setRemoteDescription("answer", answer); 2029 OffererJS->AnswererJS: ACK 2031 A.2. Example Session Descriptions 2033 A.2.1. createOffer 2035 This SDP shows a typical initial offer, created by createOffer for a 2036 PeerConnection with a single audio MediaStreamTrack, a single video 2037 MediaStreamTrack, and a single data channel. Host candidates have 2038 also already been gathered. Note some lines have been broken into 2039 two lines for formatting reasons. 2041 v=0 2042 o=- 4962303333179871722 1 IN IP4 0.0.0.0 2043 s=- 2044 t=0 0 2045 a=msid-semantic:WMS 2046 a=group:BUNDLE audio video data 2047 m=audio 56500 UDP/TLS/RTP/SAVPF 111 0 8 126 2048 c=IN IP4 192.0.2.1 2049 a=rtcp:56501 IN IP4 192.0.2.1 2050 a=candidate:3348148302 1 udp 2113937151 192.0.2.1 56500 2051 typ host generation 0 2052 a=candidate:3348148302 2 udp 2113937151 192.0.2.1 56501 2053 typ host generation 0 2054 a=ice-ufrag:ETEn1v9DoTMB9J4r 2055 a=ice-pwd:OtSK0WpNtpUjkY4+86js7ZQl 2056 a=ice-options:trickle 2057 a=mid:audio 2058 a=extmap:1 urn:ietf:params:rtp-hdrext:ssrc-audio-level 2059 a=sendrecv 2060 a=rtcp-mux 2061 a=rtcp-rsize 2062 a=fingerprint:sha-256 2063 19:E2:1C:3B:4B:9F:81:E6:B8:5C:F4:A5:A8:D8:73:04 2064 :BB:05:2F:70:9F:04:A9:0E:05:E9:26:33:E8:70:88:A2 2065 a=setup:actpass 2066 a=rtpmap:111 opus/48000/2 2067 a=fmtp:111 minptime=10 2068 a=rtpmap:0 PCMU/8000 2069 a=rtpmap:8 PCMA/8000 2070 a=rtpmap:126 telephone-event/8000 2071 a=maxptime:60 2072 a=ssrc:1732846380 cname:EocUG1f0fcg/yvY7 2073 a=msid:47017fee-b6c1-4162-929c-a25110252400 2074 f83006c5-a0ff-4e0a-9ed9-d3e6747be7d9 2075 m=video 56502 UDP/TLS/RTP/SAVPF 100 115 116 117 2076 c=IN IP4 192.0.2.1 2077 a=rtcp:56503 IN IP4 192.0.2.1 2078 a=candidate:3348148302 1 udp 2113937151 192.0.2.1 56502 2079 typ host generation 0 2080 a=candidate:3348148302 2 udp 2113937151 192.0.2.1 56503 2081 typ host generation 0 2082 a=ice-ufrag:BGKkWnG5GmiUpdIV 2083 a=ice-pwd:mqyWsAjvtKwTGnvhPztQ9mIf 2084 a=ice-options:trickle 2085 a=mid:video 2086 a=extmap:2 urn:ietf:params:rtp-hdrext:toffset 2087 a=extmap:3 http://www.webrtc.org/experiments/rtp-hdrext/abs-send-time 2088 a=sendrecv 2089 a=rtcp-mux 2090 a=rtcp-rsize 2091 a=fingerprint:sha-256 2092 19:E2:1C:3B:4B:9F:81:E6:B8:5C:F4:A5:A8:D8:73:04 2093 :BB:05:2F:70:9F:04:A9:0E:05:E9:26:33:E8:70:88:A2 2094 a=setup:actpass 2095 a=rtpmap:100 VP8/90000 2096 a=rtcp-fb:100 ccm fir 2097 a=rtcp-fb:100 nack 2098 a=rtcp-fb:100 goog-remb 2099 a=rtpmap:115 rtx/90000 2100 a=fmtp:115 apt=100 2101 a=rtpmap:116 red/90000 2102 a=rtpmap:117 ulpfec/90000 2103 a=ssrc:1366781083 cname:EocUG1f0fcg/yvY7 2104 a=ssrc:1366781084 cname:EocUG1f0fcg/yvY7 2105 a=ssrc:1366781085 cname:EocUG1f0fcg/yvY7 2106 a=ssrc-group:FID 1366781083 1366781084 2107 a=ssrc-group:FEC 1366781083 1366781085 2108 a=msid:61317484-2ed4-49d7-9eb7-1414322a7aae 2109 f30bdb4a-5db8-49b5-bcdc-e0c9a23172e0 2110 m=application 56504 DTLS/SCTP 5000 2111 c=IN IP4 192.0.2.1 2112 a=candidate:3348148302 1 udp 2113937151 192.0.2.1 56504 2113 typ host generation 0 2114 a=ice-ufrag:VD5v2BnbZm3mgP3d 2115 a=ice-pwd:+Jlkuox+VVIUDqxcfIDuTZMH 2116 a=ice-options:trickle 2117 a=mid:data 2118 a=fingerprint:sha-256 19:E2:1C:3B:4B:9F:81:E6:B8:5C:F4:A5:A8:D8:73:04 2119 :BB:05:2F:70:9F:04:A9:0E:05:E9:26:33:E8:70:88:A2 2120 a=setup:actpass 2121 a=sctpmap:5000 webrtc-datachannel 16 2123 A.2.2. createAnswer 2125 This SDP shows a typical initial answer to the above offer, created 2126 by createAnswer for a PeerConnection with a single audio 2127 MediaStreamTrack, a single video MediaStreamTrack, and a single data 2128 channel. Host candidates have also already been gathered. Note some 2129 lines have been broken into two lines for formatting reasons. 2131 v=0 2132 o=- 6729291447651054566 1 IN IP4 0.0.0.0 2133 s=- 2134 t=0 0 2135 a=msid-semantic:WMS 2136 a=group:BUNDLE audio video data 2137 m=audio 20000 UDP/TLS/RTP/SAVPF 111 0 8 126 2138 c=IN IP4 192.0.2.2 2139 a=candidate:2299743422 1 udp 2113937151 192.0.2.2 20000 2140 typ host generation 0 2141 a=ice-ufrag:6sFvz2gdLkEwjZEr 2142 a=ice-pwd:cOTZKZNVlO9RSGsEGM63JXT2 2143 a=fingerprint:sha-256 6B:8B:F0:65:5F:78:E2:51:3B:AC:6F:F3:3F:46:1B:35 2144 :DC:B8:5F:64:1A:24:C2:43:F0:A1:58:D0:A1:2C:19:08 2145 a=setup:active 2146 a=mid:audio 2147 a=extmap:1 urn:ietf:params:rtp-hdrext:ssrc-audio-level 2148 a=sendrecv 2149 a=rtcp-mux 2150 a=rtpmap:111 opus/48000/2 2151 a=fmtp:111 minptime=10 2152 a=rtpmap:0 PCMU/8000 2153 a=rtpmap:8 PCMA/8000 2154 a=rtpmap:126 telephone-event/8000 2155 a=maxptime:60 2156 a=ssrc:3429951804 cname:Q/NWs1ao1HmN4Xa5 2157 a=msid:PI39StLS8W7ZbQl1sJsWUXkr3Zf12fJUvzQ1 2158 PI39StLS8W7ZbQl1sJsWUXkr3Zf12fJUvzQ1a0 2159 m=video 20000 UDP/TLS/RTP/SAVPF 100 115 116 117 2160 c=IN IP4 192.0.2.2 2161 a=candidate:2299743422 1 udp 2113937151 192.0.2.2 20000 2162 typ host generation 0 2163 a=ice-ufrag:6sFvz2gdLkEwjZEr 2164 a=ice-pwd:cOTZKZNVlO9RSGsEGM63JXT2 2165 a=fingerprint:sha-256 6B:8B:F0:65:5F:78:E2:51:3B:AC:6F:F3:3F:46:1B:35 2166 :DC:B8:5F:64:1A:24:C2:43:F0:A1:58:D0:A1:2C:19:08 2167 a=setup:active 2168 a=mid:video 2169 a=extmap:2 urn:ietf:params:rtp-hdrext:toffset 2170 a=extmap:3 http://www.webrtc.org/experiments/rtp-hdrext/abs-send-time 2171 a=sendrecv 2172 a=rtcp-mux 2173 a=rtpmap:100 VP8/90000 2174 a=rtcp-fb:100 ccm fir 2175 a=rtcp-fb:100 nack 2176 a=rtcp-fb:100 goog-remb 2177 a=rtpmap:115 rtx/90000 2178 a=fmtp:115 apt=100 2179 a=rtpmap:116 red/90000 2180 a=rtpmap:117 ulpfec/90000 2181 a=ssrc:3229706345 cname:Q/NWs1ao1HmN4Xa5 2182 a=ssrc:3229706346 cname:Q/NWs1ao1HmN4Xa5 2183 a=ssrc:3229706347 cname:Q/NWs1ao1HmN4Xa5 2184 a=ssrc-group:FID 3229706345 3229706346 2185 a=ssrc-group:FEC 3229706345 3229706347 2186 a=msid:PI39StLS8W7ZbQl1sJsWUXkr3Zf12fJUvzQ1 2187 PI39StLS8W7ZbQl1sJsWUXkr3Zf12fJUvzQ1v0 2188 m=application 20000 DTLS/SCTP 5000 2189 c=IN IP4 192.0.2.2 2190 a=candidate:2299743422 1 udp 2113937151 192.0.2.2 20000 2191 typ host generation 0 2192 a=ice-ufrag:6sFvz2gdLkEwjZEr 2193 a=ice-pwd:cOTZKZNVlO9RSGsEGM63JXT2 2194 a=fingerprint:sha-256 6B:8B:F0:65:5F:78:E2:51:3B:AC:6F:F3:3F:46:1B:35 2195 :DC:B8:5F:64:1A:24:C2:43:F0:A1:58:D0:A1:2C:19:08 2196 a=setup:active 2197 a=mid:data 2198 a=sctpmap:5000 webrtc-datachannel 16 2200 A.2.3. Call Flows 2202 Example SDP for WebRTC call flows can be found in 2203 [I-D.nandakumar-rtcweb-sdp]. [TODO: should these call flows be 2204 merged into this section?] 2206 Appendix B. Change log 2208 Changes in draft-06: 2209 o Reworked handling of m= line recycling. 2210 o Added handling of BUNDLE and bundle-only. 2211 o Clarified handling of rollback. 2212 o Added text describing the ICE Candidate Pool and its behavior. 2213 o Allowed OfferToReceiveX to create multiple recvonly m= sections. 2215 Changes in draft-05: 2216 o Fixed several issues identified in the createOffer/Answer sections 2217 during document review. 2218 o Updated references. 2220 Changes in draft-04: 2221 o Filled in sections on createOffer and createAnswer. 2222 o Added SDP examples. 2223 o Fixed references. 2225 Changes in draft-03: 2226 o Added text describing relationship to W3C specification 2228 Changes in draft-02: 2229 o Converted from nroff 2230 o Removed comparisons to old approaches abandoned by the working 2231 group 2232 o Removed stuff that has moved to W3C specification 2233 o Align SDP handling with W3C draft 2234 o Clarified section on forking. 2236 Changes in draft-01: 2237 o Added diagrams for architecture and state machine. 2238 o Added sections on forking and rehydration. 2239 o Clarified meaning of "pranswer" and "answer". 2240 o Reworked how ICE restarts and media directions are controlled. 2241 o Added list of parameters that can be changed in a description. 2242 o Updated suggested API and examples to match latest thinking. 2243 o Suggested API and examples have been moved to an appendix. 2245 Changes in draft -00: 2247 o Migrated from draft-uberti-rtcweb-jsep-02. 2249 Authors' Addresses 2251 Justin Uberti 2252 Google 2253 747 6th Ave S 2254 Kirkland, WA 98033 2255 USA 2257 Email: justin@uberti.name 2259 Cullen Jennings 2260 Cisco 2261 170 West Tasman Drive 2262 San Jose, CA 95134 2263 USA 2265 Email: fluffy@iii.ca 2267 Eric Rescorla (editor) 2268 Mozilla 2269 331 Evelyn Ave 2270 Mountain View, CA 94041 2271 USA 2273 Email: ekr@rtfm.com