JavaScript Session Establishment
ProtocolGoogle747 6th St SKirklandWA98033USAjustin@uberti.nameCisco400 3rd Avenue SWCalgaryABT2P 4H2Canadafluffy@iii.caMozilla331 Evelyn AveMountain ViewCA94041USAekr@rtfm.com
RAI
This document describes the mechanisms for allowing a
JavaScript application to control the signaling plane of a
multimedia session via the interface specified in the W3C
RTCPeerConnection API, and discusses how this relates to existing
signaling protocols.This document describes how the W3C WEBRTC RTCPeerConnection
interface
is used to control the setup,
management and teardown of a multimedia session.WebRTC call setup has been designed to focus on controlling
the media plane, leaving signaling plane behavior up to the
application as much as possible. The rationale is that
different applications may prefer to use different protocols,
such as the existing SIP call signaling protocol, or something
custom to the particular application, perhaps for a novel use
case. In this approach, the key information that needs to be
exchanged is the multimedia session description, which
specifies the necessary transport and media configuration
information necessary to establish the media plane.With these considerations in mind, this document describes
the JavaScript Session Establishment Protocol (JSEP) that
allows for full control of the signaling state machine from
JavaScript. As described above, JSEP assumes a model in which a
JavaScript application executes inside a runtime containing
WebRTC APIs (the "JSEP implementation"). The JSEP
implementation is almost entirely divorced from the core
signaling flow, which is instead handled by the JavaScript
making use of two interfaces: (1) passing in local and remote
session descriptions and (2) interacting with the ICE state
machine. The combination of the JSEP implementation and the
JavaScript application is referred to throughout this document
as a "JSEP endpoint".In this document, the use of JSEP is described as if it
always occurs between two JSEP endpoints. Note though in many
cases it will actually be between a JSEP endpoint and some kind
of server, such as a gateway or MCU. This distinction is
invisible to the JSEP endpoint; it just follows the
instructions it is given via the API.JSEP's handling of session descriptions is simple and
straightforward. Whenever an offer/answer exchange is needed,
the initiating side creates an offer by calling a createOffer()
API. The application then uses that offer to set up its local
config via the setLocalDescription() API. The offer is finally
sent off to the remote side over its preferred signaling
mechanism (e.g., WebSockets); upon receipt of that offer, the
remote party installs it using the setRemoteDescription()
API.To complete the offer/answer exchange, the remote party uses
the createAnswer() API to generate an appropriate answer,
applies it using the setLocalDescription() API, and sends the
answer back to the initiator over the signaling channel. When
the initiator gets that answer, it installs it using the
setRemoteDescription() API, and initial setup is complete. This
process can be repeated for additional offer/answer
exchanges.Regarding ICE
, JSEP decouples the ICE state
machine from the overall signaling state machine, as the ICE
state machine must remain in the JSEP implementation, because
only the implementation has the necessary knowledge of
candidates and other transport information. Performing this
separation provides additional flexibility in protocols that
decouple session descriptions from transport. For instance, in
traditional SIP, each offer or answer is self-contained,
including both the session descriptions and the transport
information. However,
allows SIP to
be used with trickle ICE
, in which the session
description can be sent immediately and the transport
information can be sent when available. Sending transport
information separately can allow for faster ICE and DTLS
startup, since ICE checks can start as soon as any transport
information is available rather than waiting for all of it.
JSEP's decoupling of the ICE and signaling state machines
allows it to accommodate either model.Through its abstraction of signaling, the JSEP approach does
require the application to be aware of the signaling process.
While the application does not need to understand the contents
of session descriptions to set up a call, the application must
call the right APIs at the right times, convert the session
descriptions and ICE information into the defined messages of
its chosen signaling protocol, and perform the reverse
conversion on the messages it receives from the other side.One way to make life easier for the application is to
provide a JavaScript library that hides this complexity from
the developer; said library would implement a given signaling
protocol along with its state machine and serialization code,
presenting a higher level call-oriented interface to the
application developer. For example, libraries exist to adapt
the JSEP API into an API suitable for a SIP or XMPP. Thus, JSEP
provides greater control for the experienced developer without
forcing any additional complexity on the novice developer.One approach that was considered instead of JSEP was to
include a lightweight signaling protocol. Instead of providing
session descriptions to the API, the API would produce and
consume messages from this protocol. While providing a more
high-level API, this put more control of signaling within the
JSEP implementation, forcing it to have to understand and
handle concepts like signaling glare (see
, Section 4).A second approach that was considered but not chosen was to
decouple the management of the media control objects from
session descriptions, instead offering APIs that would control
each component directly. This was rejected based on the
argument that requiring exposure of this level of complexity to
the application programmer would not be beneficial; it would
result in an API where even a simple example would require a
significant amount of code to orchestrate all the needed
interactions, as well as creating a large API surface that
needed to be agreed upon and documented. In addition, these API
points could be called in any order, resulting in a more
complex set of interactions with the media subsystem than the
JSEP approach, which specifies how session descriptions are to
be evaluated and applied.One variation on JSEP that was considered was to keep the
basic session description-oriented API, but to move the
mechanism for generating offers and answers out of the JSEP
implementation. Instead of providing createOffer/createAnswer
methods within the implementation, this approach would instead
expose a getCapabilities API which would provide the
application with the information it needed in order to generate
its own session descriptions. This increases the amount of work
that the application needs to do; it needs to know how to
generate session descriptions from capabilities, and especially
how to generate the correct answer from an arbitrary offer and
the supported capabilities. While this could certainly be
addressed by using a library like the one mentioned above, it
basically forces the use of said library even for a simple
example. Providing createOffer/createAnswer avoids this
problem.The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in
.JSEP does not specify a particular signaling model or state
machine, other than the generic need to exchange session
descriptions in the fashion described by
(offer/answer) in order for both
sides of the session to know how to conduct the session. JSEP
provides mechanisms to create offers and answers, as well as to
apply them to a session. However, the JSEP implementation is
totally decoupled from the actual mechanism by which these
offers and answers are communicated to the remote side,
including addressing, retransmission, forking, and glare
handling. These issues are left entirely up to the application;
the application has complete control over which offers and
answers get handed to the implementation, and when.In order to establish the media plane, the JSEP
implementation needs specific parameters to indicate what to
transmit to the remote side, as well as how to handle the media
that is received. These parameters are determined by the
exchange of session descriptions in offers and answers, and
there are certain details to this process that must be handled
in the JSEP APIs.Whether a session description applies to the local side or
the remote side affects the meaning of that description. For
example, the list of codecs sent to a remote party indicates
what the local side is willing to receive, which, when
intersected with the set of codecs the remote side supports,
specifies what the remote side should send. However, not all
parameters follow this rule; some parameters are declarative
and the remote side MUST either accept them or reject them
altogether. An example of such a parameter is the DTLS
fingerprints
, which are calculated based on
the local certificate(s) offered, and are not subject to
negotiation.In addition, various RFCs put different conditions on the
format of offers versus answers. For example, an offer may
propose an arbitrary number of m= sections (i.e., media
descriptions as described in
, Section 5.14), but an answer must
contain the exact same number as the offer.Lastly, while the exact media parameters are only known only
after an offer and an answer have been exchanged, the offerer
may receive ICE checks, and possibly media (e.g., in the case
of a re-offer after a connection has been established) before
it receives an answer. To properly process incoming media in
this case, the offerer's media handler must be aware of the
details of the offer before the answer arrives.Therefore, in order to handle session descriptions properly,
the JSEP implementation needs:
To know if a session description pertains to the local or
remote side.To know if a session description is an offer or an
answer.To allow the offer to be specified independently of the
answer.JSEP addresses this by adding both setLocalDescription
and setRemoteDescription methods and having session description
objects contain a type field indicating the type of session
description being supplied. This satisfies the requirements
listed above for both the offerer, who first calls
setLocalDescription(sdp [offer]) and then later
setRemoteDescription(sdp [answer]), as well as for the
answerer, who first calls setRemoteDescription(sdp [offer]) and
then later setLocalDescription(sdp [answer]).During the offer/answer exchange, the outstanding offer is
considered to be "pending" at the offerer and the answerer, as
it may either be accepted or rejected. If this is a re-offer,
each side will also have "current" local and remote
descriptions, which reflect the result of the last offer/answer
exchange. Sections
,
,
, and
, provide more
detail on pending and current descriptions.JSEP also allows for an answer to be treated as provisional
by the application. Provisional answers provide a way for an
answerer to communicate initial session parameters back to the
offerer, in order to allow the session to begin, while allowing
a final answer to be specified later. This concept of a final
answer is important to the offer/answer model; when such an
answer is received, any extra resources allocated by the caller
can be released, now that the exact session configuration is
known. These "resources" can include things like extra ICE
components, TURN candidates, or video decoders. Provisional
answers, on the other hand, do no such deallocation; as a
result, multiple dissimilar provisional answers, with their own
codec choices, transport parameters, etc., can be received and
applied during call setup. Note that the final answer itself
may be different than any received provisional answers.In
, the constraint at the signaling
level is that only one offer can be outstanding for a given
session, but at the media stack level, a new offer can be
generated at any point. For example, when using SIP for
signaling, if one offer is sent, then cancelled using a SIP
CANCEL, another offer can be generated even though no answer
was received for the first offer. To support this, the JSEP
media layer can provide an offer via the createOffer() method
whenever the JavaScript application needs one for the
signaling. The answerer can send back zero or more provisional
answers, and finally end the offer-answer exchange by sending a
final answer. The state machine for this is as follows:Aside from these state transitions there is no other
difference between the handling of provisional ("pranswer") and
final ("answer") answers.JSEP's session descriptions use SDP syntax for their
internal representation. While this format is not optimal for
manipulation from JavaScript, it is widely accepted, and
frequently updated with new features; any alternate encoding of
session descriptions would have to keep pace with the changes
to SDP, at least until the time that this new encoding eclipsed
SDP in popularity.However, to provide for future flexibility, the SDP syntax
is encapsulated within a SessionDescription object, which can
be constructed from SDP, and be serialized out to SDP. If
future specifications agree on a JSON format for session
descriptions, we could easily enable this object to generate
and consume that JSON.As detailed below, most applications should be able to treat
the SessionDescriptions produced and consumed by these various
API calls as opaque blobs; that is, the application will not
need to read or change them.In order to give the application control over various common
session parameters, JSEP provides control surfaces which tell
the JSEP implementation how to generate session descriptions.
This avoids the need for JavaScript to modify session
descriptions in most cases.Changes to these objects result in changes to the session
descriptions generated by subsequent createOffer/Answer
calls.RtpTransceivers allow the application to control the RTP
media associated with one m= section. Each RtpTransceiver has
an RtpSender and an RtpReceiver, which an application can use
to control the sending and receiving of RTP media. The
application may also modify the RtpTransceiver directly, for
instance, by stopping it.RtpTransceivers generally have a 1:1 mapping with m=
sections, although there may be more RtpTransceivers than m=
sections when RtpTransceivers are created but not yet
associated with a m= section, or if RtpTransceivers have been
stopped and disassociated from m= sections. An RtpTransceiver
is said to be associated with an m= section if its mid
property is non-null; otherwise it is said to be
disassociated. The associated m= section is determined using
a mapping between transceivers and m= section indices, formed
when creating an offer or applying a remote offer.An RtpTransceiver is never associated with more than one
m= section, and once a session description is applied, a m=
section is always associated with exactly one RtpTransceiver.
However, in certain cases where a m= section has been
rejected, as discussed in
below, that m= section
will be "recycled" and associated with a new RtpTransceiver
with a new mid value.RtpTransceivers can be created explicitly by the
application or implicitly by calling setRemoteDescription
with an offer that adds new m= sections.RtpSenders allow the application to control how RTP media
is sent. An RtpSender is conceptually responsible for the
outgoing RTP stream(s) described by an m= section. This
includes encoding the attached MediaStreamTrack, sending RTP
media packets, and generating/processing RTCP for the
outgoing RTP streams(s).RtpReceivers allow the application to inspect how RTP
media is received. An RtpReceiver is conceptually responsible
for the incoming RTP stream(s) described by an m= section.
This includes processing received RTP media packets, decoding
the incoming stream(s) to produce a remote MediaStreamTrack,
and generating/processing RTCP for the incoming RTP
stream(s).JSEP gathers ICE candidates as needed by the application.
Collection of ICE candidates is referred to as a gathering
phase, and this is triggered either by the addition of a new
or recycled m= section to the local session description, or
new ICE credentials in the description, indicating an ICE
restart. Use of new ICE credentials can be triggered
explicitly by the application, or implicitly by the JSEP
implementation in response to changes in the ICE
configuration.When the ICE configuration changes in a way that requires
a new gathering phase, a 'needs-ice-restart' bit is set. When
this bit is set, calls to the createOffer API will generate
new ICE credentials. This bit is cleared by a call to the
setLocalDescription API with new ICE credentials from either
an offer or an answer, i.e., from either a local- or
remote-initiated ICE restart.When a new gathering phase starts, the ICE agent will
notify the application that gathering is occurring through an
event. Then, when each new ICE candidate becomes available,
the ICE agent will supply it to the application via an
additional event; these candidates will also automatically be
added to the current and/or pending local session
description. Finally, when all candidates have been gathered,
an event will be dispatched to signal that the gathering
process is complete.Note that gathering phases only gather the candidates
needed by new/recycled/restarting m= sections; other m=
sections continue to use their existing candidates. Also, if
an m= section is bundled (either by a successful bundle
negotiation or by being marked as bundle-only), then
candidates will be gathered and exchanged for that m= section
if and only if its MID is a BUNDLE-tag, as described in
.Candidate trickling is a technique through which a caller
may incrementally provide candidates to the callee after the
initial offer has been dispatched; the semantics of "Trickle
ICE" are defined in
. This process
allows the callee to begin acting upon the call and setting
up the ICE (and perhaps DTLS) connections immediately,
without having to wait for the caller to gather all possible
candidates. This results in faster media setup in cases where
gathering is not performed prior to initiating the call.JSEP supports optional candidate trickling by providing
APIs, as described above, that provide control and feedback
on the ICE candidate gathering process. Applications that
support candidate trickling can send the initial offer
immediately and send individual candidates when they get the
notified of a new candidate; applications that do not support
this feature can simply wait for the indication that
gathering is complete, and then create and send their offer,
with all the candidates, at this time.Upon receipt of trickled candidates, the receiving
application will supply them to its ICE agent. This triggers
the ICE agent to start using the new remote candidates for
connectivity checks.In JSEP, ICE candidates are abstracted by an
IceCandidate object, and as with session descriptions, SDP
syntax is used for the internal representation.The candidate details are specified in an IceCandidate
field, using the same SDP syntax as the
"candidate-attribute" field defined in
, Section 15.1. Note that this
field does not contain an "a=" prefix, as indicated in the
following example:The IceCandidate object contains a field to indicate
which ICE ufrag it is associated with, as defined in
, Section 15.4. This value is used
to determine which session description (and thereby which
gathering phase) this IceCandidate belongs to, which helps
resolve ambiguities during ICE restarts. If this field is
absent in a received IceCandidate (perhaps when
communicating with a non-JSEP endpoint), the most recently
received session description is assumed.The IceCandidate object also contains fields to indicate
which m= section it is associated with, which can be
identified in one of two ways, either by a m= section
index, or a MID. The m= section index is a zero-based
index, with index N referring to the N+1th m= section in
the session description referenced by this IceCandidate.
The MID is a "media stream identification" value, as
defined in
, Section 4, which provides a
more robust way to identify the m= section in the session
description, using the MID of the associated RtpTransceiver
object (which may have been locally generated by the
answerer when interacting with a non-JSEP endpoint that
does not support the MID attribute, as discussed in
below). If the
MID field is present in a received IceCandidate, it MUST be
used for identification; otherwise, the m= section index is
used instead.When creating an IceCandidate object, JSEP
implementations MUST populate each of the candidate, ufrag,
m= section index, and MID fields. Implementations MUST also
be prepared to receive objects with some fields missing, as
mentioned above.Typically, when gathering ICE candidates, the JSEP
implementation will gather all possible forms of initial
candidates - host, server reflexive, and relay. However, in
certain cases, applications may want to have more specific
control over the gathering process, due to privacy or related
concerns. For example, one may want to only use relay
candidates, to leak as little location information as
possible (keeping in mind that this choice comes with
corresponding operational costs). To accomplish this, JSEP
allows the application to restrict which ICE candidates are
used in a session. Note that this filtering is applied on top
of any restrictions the implementation chooses to enforce
regarding which IP addresses are permitted for the
application, as discussed in
.There may also be cases where the application wants to
change which types of candidates are used while the session
is active. A prime example is where a callee may initially
want to use only relay candidates, to avoid leaking location
information to an arbitrary caller, but then change to use
all candidates (for lower operational cost) once the user has
indicated they want to take the call. For this scenario, the
JSEP implementation MUST allow the candidate policy to be
changed in mid-session, subject to the aforementioned
interactions with local policy.To administer the ICE candidate policy, the JSEP
implementation will determine the current setting at the
start of each gathering phase. Then, during the gathering
phase, the implementation MUST NOT expose candidates
disallowed by the current policy to the application, use them
as the source of connectivity checks, or indirectly expose
them via other fields, such as the raddr/rport attributes for
other ICE candidates. Later, if a different policy is
specified by the application, the application can apply it by
kicking off a new gathering phase via an ICE restart.JSEP applications typically inform the JSEP implementation
to begin ICE gathering via the information supplied to
setLocalDescription, as the local description indicates the
number of ICE components which will be needed and for which
candidates must be gathered. However, to accelerate cases
where the application knows the number of ICE components to
use ahead of time, it may ask the implementation to gather a
pool of potential ICE candidates to help ensure rapid media
setup.When setLocalDescription is eventually called, and the
JSEP implementation goes to gather the needed ICE candidates,
it SHOULD start by checking if any candidates are available
in the pool. If there are candidates in the pool, they SHOULD
be handed to the application immediately via the ICE
candidate event. If the pool becomes depleted, either because
a larger-than-expected number of ICE components is used, or
because the pool has not had enough time to gather
candidates, the remaining candidates are gathered as usual.
This only occurs for the first offer/answer exchange, after
which the candidate pool is emptied and no longer used.One example of where this concept is useful is an
application that expects an incoming call at some point in
the future, and wants to minimize the time it takes to
establish connectivity, to avoid clipping of initial media.
By pre-gathering candidates into the pool, it can exchange
and start sending connectivity checks from these candidates
almost immediately upon receipt of a call. Note though that
by holding on to these pre-gathered candidates, which will be
kept alive as long as they may be needed, the application
will consume resources on the STUN/TURN servers it is
using.Video size negotiation is the process through which a
receiver can use the "a=imageattr" SDP attribute
to indicate what video frame sizes it
is capable of receiving. A receiver may have hard limits on
what its video decoder can process, or it may have some maximum
set by policy. By specifying these limits in an "a=imageattr"
attribute, JSEP endpoints can attempt to ensure that the remote
sender transmits video at an acceptable resolution. However,
when communicating with a non-JSEP endpoint that does not
understand this attribute, any signaled limits may be exceeded,
and the JSEP implementation MUST handle this gracefully, e.g.,
by discarding the video.Note that certain codecs support transmission of samples
with aspect ratios other than 1.0 (i.e., non-square pixels).
JSEP implementations will not transmit non-square pixels, but
SHOULD receive and render such video with the correct aspect
ratio. However, sample aspect ratio has no impact on the size
negotiation described below; all dimensions are measured in
pixels, whether square or not.The receiver will first intersect any known local limits
(e.g., hardware decoder capababilities, local policy) to
determine the absolute minimum and maximum sizes it can
receive. If there are no known local limits, the
"a=imageattr" attribute SHOULD be omitted. If these local
limits preclude receiving any video, i.e., the degenerate
case of no permitted resolutions, the "a=imageattr" attribute
MUST be omitted, and the m= section MUST be marked as
sendonly/inactive, as appropriate.Otherwise, an "a=imageattr" attribute is created with
"recv" direction, and the resulting resolution space formed
from the aforementioned intersection is used to specify its
minimum and maximum x= and y= values.The rules here express a single set of preferences, and
therefore, the "a=imageattr" q= value is not important. It
SHOULD be set to 1.0.The "a=imageattr" field is payload type specific. When all
video codecs supported have the same capabilities, use of a
single attribute, with the wildcard payload type (*), is
RECOMMENDED. However, when the supported video codecs have
different limitations, specific "a=imageattr" attributes MUST
be inserted for each payload type.As an example, consider a system with a multiformat video
decoder, which is capable of decoding any resolution from
48x48 to 720p, In this case, the implementation would
generate this attribute:a=imageattr:* recv [x=[48:1280],y=[48:720],q=1.0]This declaration indicates that the receiver is capable of
decoding any image resolution from 48x48 up to 1280x720
pixels. defines "a=imageattr" to be an
advisory field. This means that it does not absolutely
constrain the video formats that the sender can use, but
gives an indication of the preferred values.This specification prescribes more specific behavior. When
a MediaStreamTrack, which is producing video of a certain
resolution (the "track resolution"), is attached to a
RtpSender, which is encoding the track video at the same or
lower resolution(s) (the "encoder resolutions"), and a remote
description is applied that references the sender and
contains valid "a=imageattr recv" attributes, it MUST follow
the rules below to ensure the sender does not transmit a
resolution that would exceed the size criteria specified in
the attributes. These rules MUST be followed as long as the
attributes remain present in the remote description,
including cases in which the track changes its resolution, or
is replaced with a different track.Depending on how the RtpSender is configured, it may be
producing a single encoding at a certain resolution, or, if
simulcast
has been negotiated, multiple
encodings, each at their own specific resolution. In
addition, depending on the configuration, each encoding may
have the flexibility to reduce resolution when needed, or may
be locked to a specific output resolution.For each encoding being produced by the RtpSender, the set
of "a=imageattr recv" attributes in the corresponding m=
section of the remote description is processed to determine
what should be transmitted. Only attributes that reference
the media format selected for the encoding are considered;
each such attribute is evaluated individually, starting with
the attribute with the highest "q=" value. If multiple
attributes have the same "q=" value, they are evaluated in
the order they appear in their containing m= section. Note
that while JSEP endpoints will include at most one
"a=imageattr recv" attribute per media format, JSEP endpoints
may receive session descriptions from non-JSEP endpoints with
m= sections that contain multiple such attributes.For each "a=imageattr recv" attribute, the following rules
are applied. If this processing is successful, the encoding
is transmitted accordingly, and no further attributes are
considered for that encoding. Otherwise, the next attribute
is evaluated, in the aforementioned order. If none of the
supplied attributes can be processed successfully, the
encoding MUST NOT be transmitted, and an error SHOULD be
raised to the application.
The limits from the attribute are compared to the
encoder resolution. Only the specific limits mentioned
below are considered; any other values, such as picture
aspect ratio, MUST be ignored. When considering a
MediaStreamTrack that is producing rotated video, the
unrotated resolution MUST be used for the checks. This is
required regardless of whether the receiver supports
performing receive-side rotation (e.g., through CVO
), as it significantly simplifies
the matching logic.If the attribute includes a "sar=" (sample aspect ratio)
value set to something other than "1.0", indicating the
receiver wants to receive non-square pixels, this cannot be
satisfied and the attribute MUST NOT be used.If the encoder resolution exceeds the maximum size
permitted by the attribute, and the encoder is allowed to
adjust its resolution, the encoder SHOULD apply downscaling
in order to satisfy the limits. Downscaling MUST NOT change
the picture aspect ratio of the encoding, ignoring any
trivial differences due to rounding. For example, if the
encoder resolution is 1280x720, and the attribute specified
a maximum of 640x480, the expected output resolution would
be 640x360. If downscaling cannot be applied, the attribute
MUST NOT be used.If the encoder resolution is less than the minimum size
permitted by the attribute, the attribute MUST NOT be used;
the encoder MUST NOT apply upscaling. JSEP implementations
SHOULD avoid this situation by allowing receipt of
arbitrarily small resolutions, perhaps via fallback to a
software decoder.If the encoder resolution is within the maximum and
minimum sizes, no action is needed.JSEP supports simulcast transmission of a MediaStreamTrack,
where multiple encodings of the source media can be transmitted
within the context of a single m= section. The current JSEP API
is designed to allow applications to send simulcasted media but
only to receive a single encoding. This allows for multi-user
scenarios where each sending client sends multiple encodings to
a server, which then, for each receiving client, chooses the
appropriate encoding to forward.Applications request support for simulcast by configuring
multiple encodings on an RtpSender. Upon generation of an offer
or answer, these encodings are indicated via SDP markings on
the corresponding m= section, as described below. Receivers
that understand simulcast and are willing to receive it will
also include SDP markings to indicate their support, and JSEP
endpoints will use these markings to determine whether
simulcast is permitted for a given RtpSender. If simulcast
support is not negotiated, the RtpSender will only use the
first configured encoding.Note that the exact simulcast parameters are up to the
sending application. While the aforementioned SDP markings are
provided to ensure the remote side can receive and demux
multiple simulcast encodings, the specific resolutions and
bitrates to be used for each encoding are purely a send-side
decision in JSEP.JSEP currently does not provide a mechanism to configure
receipt of simulcast. This means that if simulcast is offered
by the remote endpoint, the answer generated by a JSEP endpoint
will not indicate support for receipt of simulcast, and as such
the remote endpoint will only send a single encoding per m=
section.In addition, JSEP does not provide a mechanism to handle an
incoming offer requesting simulcast from the JSEP endpoint.
This means that setting up simulcast in the case where the JSEP
endpoint receives the initial offer requires out-of-band
signaling or SDP inspection. However, in the case where the
JSEP endpoint sets up simulcast in its in initial offer, any
established simulcast streams will continue to work upon
receipt of an incoming re-offer. Future versions of this
specification may add additional APIs to handle the incoming
initial offer scenario.When using JSEP to transmit multiple encodings from a
RtpSender, the techniques from
and
are used. Specifically,
when multiple encodings have been configured for a RtpSender,
the m= section for the RtpSender will include an "a=simulcast"
attribute, as defined in
, Section 6.2,
with a "send" simulcast stream description that lists each
desired encoding, and no "recv" simulcast stream description.
The m= section will also include an "a=rid" attribute for each
encoding, as specified in
, Section 4; the use of
RID identifiers allows the individual encodings to be
disambiguated even though they are all part of the same m=
section.Some call signaling systems allow various types of forking
where an SDP Offer may be provided to more than one device. For
example, SIP
defines both a "Parallel Search"
and "Sequential Search". Although these are primarily signaling
level issues that are outside the scope of JSEP, they do have
some impact on the configuration of the media plane that is
relevant. When forking happens at the signaling layer, the
JavaScript application responsible for the signaling needs to
make the decisions about what media should be sent or received
at any point of time, as well as which remote endpoint it
should communicate with; JSEP is used to make sure the media
engine can make the RTP and media perform as required by the
application. The basic operations that the applications can
have the media engine do are:
Start exchanging media with a given remote peer, but keep
all the resources reserved in the offer.Start exchanging media with a given remote peer, and free
any resources in the offer that are not being used.Sequential forking involves a call being dispatched to
multiple remote callees, where each callee can accept the
call, but only one active session ever exists at a time; no
mixing of received media is performed.JSEP handles sequential forking well, allowing the
application to easily control the policy for selecting the
desired remote endpoint. When an answer arrives from one of
the callees, the application can choose to apply it either as
a provisional answer, leaving open the possibility of using a
different answer in the future, or apply it as a final
answer, ending the setup flow.In a "first-one-wins" situation, the first answer will be
applied as a final answer, and the application will reject
any subsequent answers. In SIP parlance, this would be ACK +
BYE.In a "last-one-wins" situation, all answers would be
applied as provisional answers, and any previous call leg
will be terminated. At some point, the application will end
the setup process, perhaps with a timer; at this point, the
application could reapply the pending remote description as a
final answer.Parallel forking involves a call being dispatched to
multiple remote callees, where each callee can accept the
call, and multiple simultaneous active signaling sessions can
be established as a result. If multiple callees send media at
the same time, the possibilities for handling this are
described in
, Section 3.1. Most SIP devices
today only support exchanging media with a single device at a
time, and do not try to mix multiple early media audio
sources, as that could result in a confusing situation. For
example, consider having a European ringback tone mixed
together with the North American ringback tone - the
resulting sound would not be like either tone, and would
confuse the user. If the signaling application wishes to only
exchange media with one of the remote endpoints at a time,
then from a media engine point of view, this is exactly like
the sequential forking case.In the parallel forking case where the JavaScript
application wishes to simultaneously exchange media with
multiple peers, the flow is slightly more complex, but the
JavaScript application can follow the strategy that
describes using UPDATE. The
UPDATE approach allows the signaling to set up a separate
media flow for each peer that it wishes to exchange media
with. In JSEP, this offer used in the UPDATE would be formed
by simply creating a new PeerConnection (see
) and making sure that
the same local media streams have been added into this new
PeerConnection. Then the new PeerConnection object would
produce a SDP offer that could be used by the signaling to
perform the UPDATE strategy discussed in
.As a result of sharing the media streams, the application
will end up with N parallel PeerConnection sessions, each
with a local and remote description and their own local and
remote addresses. The media flow from these sessions can be
managed using setDirection (see
), or the
application can choose to play out the media from all
sessions mixed together. Of course, if the application wants
to only keep a single session, it can simply terminate the
sessions that it no longer needs.This section details the basic operations that must be present
to implement JSEP functionality. The actual API exposed in the
W3C API may have somewhat different syntax, but should map easily
to these concepts.The PeerConnection constructor allows the application to
specify global parameters for the media session, such as the
STUN/TURN servers and credentials to use when gathering
candidates, as well as the initial ICE candidate policy and
pool size, and also the bundle policy to use.If an ICE candidate policy is specified, it functions as
described in
, causing the JSEP
implementation to only surface the permitted candidates
(including any implementation-internal filtering) to the
application, and only use those candidates for connectivity
checks. The set of available policies is as follows:
All candidates permitted by
implementation policy will be gathered and used.All candidates except relay candidates
will be filtered out. This obfuscates the location
information that might be ascertained by the remote peer
from the received candidates. Depending on how the
application deploys and chooses relay servers, this could
obfuscate location to a metro or possibly even global
level.The default ICE candidate policy MUST be set to "all" as
this is generally the desired policy, and also typically
reduces use of application TURN server resources
significantly.If a size is specified for the ICE candidate pool, this
indicates the number of ICE components to pre-gather
candidates for. Because pre-gathering results in utilizing
STUN/TURN server resources for potentially long periods of
time, this must only occur upon application request, and
therefore the default candidate pool size MUST be zero.The application can specify its preferred policy regarding
use of bundle, the multiplexing mechanism defined in
. Regardless of policy, the application will always
try to negotiate bundle onto a single transport, and will
offer a single bundle group across all m= sections; use of
this single transport is contingent upon the answerer
accepting bundle. However, by specifying a policy from the
list below, the application can control exactly how
aggressively it will try to bundle media streams together,
which affects how it will interoperate with a
non-bundle-aware endpoint. When negotiating with a
non-bundle-aware endpoint, only the streams not marked as
bundle-only streams will be established.The set of available policies is as follows:
The first m= section of each type
(audio, video, or application) will contain transport
parameters, which will allow an answerer to unbundle that
section. The second and any subsequent m= section of each
type will be marked bundle-only. The result is that if
there are N distinct media types, then candidates will be
gathered for for N media streams. This policy balances
desire to multiplex with the need to ensure basic audio and
video can still be negotiated in legacy cases. When acting
as answerer, if there is no bundle group in the offer, the
implementation will reject all but the first m= section of
each type.All m= sections will contain
transport parameters; none will be marked as bundle-only.
This policy will allow all streams to be received by
non-bundle-aware endpoints, but require separate candidates
to be gathered for each media stream.Only the first m= section will
contain transport parameters; all streams other than the
first will be marked as bundle-only. This policy aims to
minimize candidate gathering and maximize multiplexing, at
the cost of less compatibility with legacy endpoints. When
acting as answerer, the implementation will reject any m=
sections other than the first m= section, unless they are
in the same bundle group as that m= section.As it provides the best tradeoff between performance and
compatibility with legacy endpoints, the default bundle
policy MUST be set to "balanced".The application can specify its preferred policy regarding
use of RTP/RTCP multiplexing
using one of the following
policies:
The JSEP implementation will
gather both RTP and RTCP candidates but also will offer
"a=rtcp-mux", thus allowing for compatibility with either
multiplexing or non-multiplexing endpoints.The JSEP implementation will only
gather RTP candidates and will insert an "a=rtcp-mux-only"
indication into any new m= sections in offers it generates.
This halves the number of candidates that the offerer needs
to gather. Applying a description with an m= section that
does not contain an "a=rtcp-mux" attribute will cause an
error to be returned.The default multiplexing policy MUST be set to "require".
Implementations MAY choose to reject attempts by the
application to set the multiplexing policy to
"negotiate".The addTrack method adds a MediaStreamTrack to the
PeerConnection, using the MediaStream argument to associate
the track with other tracks in the same MediaStream, so that
they can be added to the same "LS" group when creating an
offer or answer. Adding tracks to the same "LS" group
indicates that the playback of these tracks should be
synchronized for proper lip sync, as described in
, Section 7. addTrack attempts
to minimize the number of transceivers as follows: If the
PeerConnection is in the "have-remote-offer" state, the track
will be attached to the first compatible transceiver that was
created by the most recent call to setRemoteDescription() and
does not have a local track. Otherwise, a new transceiver
will be created, as described in
.The removeTrack method removes a MediaStreamTrack from the
PeerConnection, using the RtpSender argument to indicate
which sender should have its track removed. The sender's
track is cleared, and the sender stops sending. Future calls
to createOffer will mark the m= section associated with the
sender as recvonly (if transceiver.direction is sendrecv) or
as inactive (if transceiver.direction is sendonly).The addTransceiver method adds a new RtpTransceiver to the
PeerConnection. If a MediaStreamTrack argument is provided,
then the transceiver will be configured with that media type
and the track will be attached to the transceiver. Otherwise,
the application MUST explicitly specify the type; this mode
is useful for creating recvonly transceivers as well as for
creating transceivers to which a track can be attached at
some later point.At the time of creation, the application can also specify
a transceiver direction attribute, a set of MediaStreams
which the transceiver is associated with (allowing LS group
assignments), and a set of encodings for the media (used for
simulcast as described in
).The createDataChannel method creates a new data channel
and attaches it to the PeerConnection. If no data channel
currently exists for this PeerConnection, then a new
offer/answer exchange is required. All data channels on a
given PeerConnection share the same SCTP/DTLS association and
therefore the same m= section, so subsequent creation of data
channels does not have any impact on the JSEP state.The createDataChannel method also includes a number of
arguments which are used by the PeerConnection (e.g.,
maxPacketLifetime) but are not reflected in the SDP and do
not affect the JSEP state.The createOffer method generates a blob of SDP that
contains a
offer with the supported
configurations for the session, including descriptions of the
media added to this PeerConnection, the codec/RTP/RTCP
options supported by this implementation, and any candidates
that have been gathered by the ICE agent. An options
parameter may be supplied to provide additional control over
the generated offer. This options parameter allows an
application to trigger an ICE restart, for the purpose of
reestablishing connectivity.In the initial offer, the generated SDP will contain all
desired functionality for the session (functionality that is
supported but not desired by default may be omitted); for
each SDP line, the generation of the SDP will follow the
process defined for generating an initial offer from the
document that specifies the given SDP line. The exact
handling of initial offer generation is detailed in
below.In the event createOffer is called after the session is
established, createOffer will generate an offer to modify the
current session based on any changes that have been made to
the session, e.g., adding or stopping RtpTransceivers, or
requesting an ICE restart. For each existing stream, the
generation of each SDP line must follow the process defined
for generating an updated offer from the RFC that specifies
the given SDP line. For each new stream, the generation of
the SDP must follow the process of generating an initial
offer, as mentioned above. If no changes have been made, or
for SDP lines that are unaffected by the requested changes,
the offer will only contain the parameters negotiated by the
last offer-answer exchange. The exact handling of subsequent
offer generation is detailed in
. below.Session descriptions generated by createOffer must be
immediately usable by setLocalDescription; if a system has
limited resources (e.g. a finite number of decoders),
createOffer should return an offer that reflects the current
state of the system, so that setLocalDescription will succeed
when it attempts to acquire those resources.Calling this method may do things such as generating new
ICE credentials, but does not change the PeerConnection
state, trigger candidate gathering, or cause media to start
or stop flowing. Specifically, the offer is not applied, and
does not become the pending local description, until
setLocalDescription is called.The createAnswer method generates a blob of SDP that
contains a
SDP answer with the supported
configuration for the session that is compatible with the
parameters supplied in the most recent call to
setRemoteDescription, which MUST have been called prior to
calling createAnswer. Like createOffer, the returned blob
contains descriptions of the media added to this
PeerConnection, the codec/RTP/RTCP options negotiated for
this session, and any candidates that have been gathered by
the ICE agent. An options parameter may be supplied to
provide additional control over the generated answer.As an answer, the generated SDP will contain a specific
configuration that specifies how the media plane should be
established; for each SDP line, the generation of the SDP
must follow the process defined for generating an answer from
the document that specifies the given SDP line. The exact
handling of answer generation is detailed in
. below.Session descriptions generated by createAnswer must be
immediately usable by setLocalDescription; like createOffer,
the returned description should reflect the current state of
the system.Calling this method may do things such as generating new
ICE credentials, but does not change the PeerConnection
state, trigger candidate gathering, or or cause a media state
change. Specifically, the answer is not applied, and does not
become the current local description, until
setLocalDescription is called.Session description objects (RTCSessionDescription) may be
of type "offer", "pranswer", "answer" or "rollback". These
types provide information as to how the description parameter
should be parsed, and how the media state should be
changed."offer" indicates that a description should be parsed as
an offer; said description may include many possible media
configurations. A description used as an "offer" may be
applied anytime the PeerConnection is in a stable state, or
as an update to a previously supplied but unanswered
"offer"."pranswer" indicates that a description should be parsed
as an answer, but not a final answer, and so should not
result in the freeing of allocated resources. It may result
in the start of media transmission, if the answer does not
specify an inactive media direction. A description used as a
"pranswer" may be applied as a response to an "offer", or an
update to a previously sent "pranswer"."answer" indicates that a description should be parsed as
an answer, the offer-answer exchange should be considered
complete, and any resources (decoders, candidates) that are
no longer needed can be released. A description used as an
"answer" may be applied as a response to an "offer", or an
update to a previously sent "pranswer".The only difference between a provisional and final answer
is that the final answer results in the freeing of any unused
resources that were allocated as a result of the offer. As
such, the application can use some discretion on whether an
answer should be applied as provisional or final, and can
change the type of the session description as needed. For
example, in a serial forking scenario, an application may
receive multiple "final" answers, one from each remote
endpoint. The application could choose to accept the initial
answers as provisional answers, and only apply an answer as
final when it receives one that meets its criteria (e.g. a
live user instead of voicemail)."rollback" is a special session description type implying
that the state machine should be rolled back to the previous
stable state, as described in
. The contents MUST be
empty.Most applications will not need to create answers using
the "pranswer" type. While it is good practice to send an
immediate response to an offer, in order to warm up the
session transport and prevent media clipping, the preferred
handling for a JSEP application is to create and send a
"sendonly" final answer with a null MediaStreamTrack
immediately after receiving the offer, which will prevent
media from being sent by the caller, and allow media to be
sent immediately upon answer by the callee. Later, when the
callee actually accepts the call, the application can plug
in the real MediaStreamTrack and create a new "sendrecv"
offer to update the previous offer/answer pair and start
bidirectional media flow. While this could also be done
with a "sendonly" pranswer, followed by a "sendrecv"
answer, the initial pranswer leaves the offer-answer
exchange open, which means that the caller cannot send an
updated offer during this time.As an example, consider a typical JSEP application that
wants to set up audio and video as quickly as possible.
When the callee receives an offer with audio and video
MediaStreamTracks, it will send an immediate answer
accepting these tracks as sendonly (meaning that the caller
will not send the callee any media yet, and because the
callee has not yet added its own MediaStreamTracks, the
callee will not send any media either). It will then ask
the user to accept the call and acquire the needed local
tracks. Upon acceptance by the user, the application will
plug in the tracks it has acquired, which, because ICE and
DTLS handshaking have likely completed by this point, can
start transmitting immediately. The application will also
send a new offer to the remote side indicating call
acceptance and moving the audio and video to be two-way
media. A detailed example flow along these lines is shown
in
.Of course, some applications may not be able to perform
this double offer-answer exchange, particularly ones that
are attempting to gateway to legacy signaling protocols. In
these cases, pranswer can still provide the application
with a mechanism to warm up the transport.In certain situations it may be desirable to "undo" a
change made to setLocalDescription or setRemoteDescription.
Consider a case where a call is ongoing, and one side wants
to change some of the session parameters; that side
generates an updated offer and then calls
setLocalDescription. However, the remote side, either
before or after setRemoteDescription, decides it does not
want to accept the new parameters, and sends a reject
message back to the offerer. Now, the offerer, and possibly
the answerer as well, need to return to a stable state and
the previous local/remote description. To support this, we
introduce the concept of "rollback", which discards any
proposed changes to the session, returning the state
machine to the stable state. A rollback is performed by
supplying a session description of type "rollback" with
empty contents to either setLocalDescription or
setRemoteDescription.The setLocalDescription method instructs the
PeerConnection to apply the supplied session description as
its local configuration. The type field indicates whether the
description should be processed as an offer, provisional
answer, final answer, or rollback; offers and answers are
checked differently, using the various rules that exist for
each SDP line.This API changes the local media state; among other
things, it sets up local resources for receiving and decoding
media. In order to successfully handle scenarios where the
application wants to offer to change from one media format to
a different, incompatible format, the PeerConnection must be
able to simultaneously support use of both the current and
pending local descriptions (e.g., support the codecs that
exist in either description). This dual processing begins
when the PeerConnection enters the "have-local-offer" state,
and continues until setRemoteDescription is called with
either a final answer, at which point the PeerConnection can
fully adopt the pending local description, or a rollback,
which results in a revert to the current local
description.This API indirectly controls the candidate gathering
process. When a local description is supplied, and the number
of transports currently in use does not match the number of
transports needed by the local description, the
PeerConnection will create transports as needed and begin
gathering candidates for each transport, using ones from the
candidate pool if available.If setRemoteDescription was previously called with an
offer, and setLocalDescription is called with an answer
(provisional or final), and the media directions are
compatible, and media is available to send, this will result
in the starting of media transmission.The setRemoteDescription method instructs the
PeerConnection to apply the supplied session description as
the desired remote configuration. As in setLocalDescription,
the type field of the description indicates how it should be
processed.This API changes the local media state; among other
things, it sets up local resources for sending and encoding
media.If setLocalDescription was previously called with an
offer, and setRemoteDescription is called with an answer
(provisional or final), and the media directions are
compatible, and media is available to send, this will result
in the starting of media transmission.The currentLocalDescription method returns the current
negotiated local description - i.e., the local description
from the last successful offer/answer exchange - in addition
to any local candidates that have been generated by the ICE
agent since the local description was set.A null object will be returned if an offer/answer exchange
has not yet been completed.The pendingLocalDescription method returns a copy of the
local description currently in negotiation - i.e., a local
offer set without any corresponding remote answer - in
addition to any local candidates that have been generated by
the ICE agent since the local description was set.A null object will be returned if the state of the
PeerConnection is "stable" or "have-remote-offer".The currentRemoteDescription method returns a copy of the
current negotiated remote description - i.e., the remote
description from the last successful offer/answer exchange -
in addition to any remote candidates that have been supplied
via processIceMessage since the remote description was
set.A null object will be returned if an offer/answer exchange
has not yet been completed.The pendingRemoteDescription method returns a copy of the
remote description currently in negotiation - i.e., a remote
offer set without any corresponding local answer - in
addition to any remote candidates that have been supplied via
processIceMessage since the remote description was set.A null object will be returned if the state of the
PeerConnection is "stable" or "have-local-offer".The canTrickleIceCandidates property indicates whether the
remote side supports receiving trickled candidates. There are
three potential values:
No SDP has been received from the other
side, so it is not known if it can handle trickle. This is
the initial value before setRemoteDescription() is
called.SDP has been received from the other
side indicating that it can support trickle.SDP has been received from the other
side indicating that it cannot support trickle.As described in
, JSEP
implementations always provide candidates to the application
individually, consistent with what is needed for Trickle ICE.
However, applications can use the canTrickleIceCandidates
property to determine whether their peer can actually do
Trickle ICE, i.e., whether it is safe to send an initial
offer or answer followed later by candidates as they are
gathered. As "true" is the only value that definitively
indicates remote Trickle ICE support, an application which
compares canTrickleIceCandidates against "true" will by
default attempt Half Trickle on initial offers and Full
Trickle on subsequent interactions with a Trickle
ICE-compatible agent.The setConfiguration method allows the global
configuration of the PeerConnection, which was initially set
by constructor parameters, to be changed during the session.
The effects of this method call depend on when it is invoked,
and differ depending on which specific parameters are
changed:Any changes to the STUN/TURN servers to use affect the
next gathering phase. If an ICE gathering phase has
already started or completed, the 'needs-ice-restart' bit
mentioned in
will be set.
This will cause the next call to createOffer to generate
new ICE credentials, for the purpose of forcing an ICE
restart and kicking off a new gathering phase, in which
the new servers will be used. If the ICE candidate pool
has a nonzero size, and a local description has not yet
been applied, any existing candidates will be discarded,
and new candidates will be gathered from the new
servers.Any change to the ICE candidate policy affects the
next gathering phase. If an ICE gathering phase has
already started or completed, the 'needs-ice-restart' bit
will be set. Either way, changes to the policy have no
effect on the candidate pool, because pooled candidates
are not made available to the application until a
gathering phase occurs, and so any necessary filtering
can still be done on any pooled candidates.The ICE candidate pool size MUST NOT be changed after
applying a local description. If a local description has
not yet been applied, any changes to the ICE candidate
pool size take effect immediately; if increased,
additional candidates are pre-gathered; if decreased, the
now-superfluous candidates are discarded.The bundle and RTCP-multiplexing policies MUST NOT be
changed after the construction of the PeerConnection.This call may result in a change to the state of the ICE
Agent.The addIceCandidate method provides an update to the ICE
agent via an IceCandidate object
. If the
IceCandidate's candidate field is filled in, the IceCandidate
is treated as a new remote ICE candidate, which will be added
to the current and/or pending remote description according to
the rules defined for Trickle ICE. Otherwise, the
IceCandidate is treated as an end-of-candidates indication,
as defined in
.In either case, the m= section index, MID, and ufrag
fields from the supplied IceCandidate are used to determine
which m= section and ICE candidate generation the
IceCandidate belongs to, as described in
above. In the case
of an end-of-candidates indication, the absence of both the
m= section index and MID fields is interpreted to mean that
the indication applies to all m= sections in the specified
ICE candidate generation. However, if both fields are absent
for a new remote candidate, this MUST be treated as an
invalid condition, as specified below.If any IceCandidate fields contain invalid values, or an
error occurs during the processing of the IceCandidate
object, the supplied IceCandidate MUST be ignored and an
error MUST be returned.Otherwise, the new remote candidate or end-of-candidates
indication is supplied to the ICE agent. In the case of a new
remote candidate, connectivity checks will be sent to the new
candidate.The stop method stops an RtpTransceiver. This will cause
future calls to createOffer to generate a zero port for the
associated m= section. See below for more details.The stopped property indicates whether the transceiver has
been stopped, either by a call to stopTransceiver or by
applying an answer that rejects the associated m= section. In
either of these cases, it is set to "true", and otherwise
will be set to "false".A stopped RtpTransceiver does not send any outgoing RTP or
RTCP or process any incoming RTP or RTCP. It cannot be
restarted.The setDirection method sets the direction of a
transceiver, which affects the direction property of the
associated m= section on future calls to createOffer and
createAnswer. The permitted values for direction are
"recvonly", "sendrecv", "sendonly", and "inactive", mirroring
the identically-named directional attributes defined in
, Section 6.When creating offers, the transceiver direction is
directly reflected in the output, even for re-offers. When
creating answers, the transceiver direction is intersected
with the offered direction, as explained in
below.Note that while setDirection sets the direction property
of the transceiver immediately (
), this property
does not immediately affect whether the transceiver's
RtpSender will send or its RtpReceiver will receive. The
direction in effect is represented by the currentDirection
property, which is only updated when an answer is
applied.The direction property indicates the last value passed
into setDirection. If setDirection has never been called, it
is set to the direction the transceiver was initialized
with.The currentDirection property indicates the last
negotiated direction for the transceiver's associated m=
section. More specifically, it indicates the
directional attribute of the
associated m= section in the last applied answer (including
provisional answers), with "send" and "recv" directions
reversed if it was a remote answer. For example, if the
directional attribute for the associated m= section in a
remote answer is "recvonly", currentDirection is set to
"sendonly".If an answer that references this transceiver has not yet
been applied, or if the transceiver is stopped,
currentDirection is set to null.The setCodecPreferences method sets the codec preferences
of a transceiver, which in turn affect the presence and order
of codecs of the associated m= section on future calls to
createOffer and createAnswer. Note that setCodecPreferences
does not directly affect which codec the implementation
decides to send. It only affects which codecs the
implementation indicates that it prefers to receive, via the
offer or answer. Even when a codec is excluded by
setCodecPreferences, it still may be used to send until the
next offer/answer exchange discards it.The codec preferences of an RtpTransceiver can cause
codecs to be excluded by subsequent calls to createOffer and
createAnswer, in which case the corresponding media formats
in the associated m= section will be excluded. The codec
preferences cannot add media formats that would otherwise not
be present.The codec preferences of an RtpTransceiver can also
determine the order of codecs in subsequent calls to
createOffer and createAnswer, in which case the order of the
media formats in the associated m= section will follow the
specified preferences.This section describes the specific procedures to be followed
when creating and parsing SDP objects.JSEP implementations must comply with the specifications
listed below that govern the creation and processing of offers
and answers.All session descriptions handled by JSEP implementations,
both local and remote, MUST indicate support for the
following specifications. If any of these are absent, this
omission MUST be treated as an error.
ICE, as specified in
, MUST be used. Note that the
remote endpoint may use a Lite implementation;
implementations MUST properly handle remote endpoints which
do ICE-Lite.DTLS
or DTLS-SRTP
, MUST be used, as
appropriate for the media type, as specified in
The SDES SRTP keying mechanism from
MUST NOT be used, as discussed in
.For media m= sections, JSEP implementations MUST support
the "UDP/TLS/RTP/SAVPF" profile specified in
, and MUST indicate this
profile for each media m= line they produce in an offer. For
data m= sections, implementations MUST support the
"UDP/DTLS/SCTP" profile and MUST indicate this profile for
each data m= line they produce in an offer. Although these
profiles are formally associated with UDP, ICE can select
either UDP
or TCP
transport depending on network
conditions, even when advertising a UDP profile.Unfortunately, in an attempt at compatibility, some
endpoints generate other profile strings even when they mean
to support one of these profiles. For instance, an endpoint
might generate "RTP/AVP" but supply "a=fingerprint" and
"a=rtcp-fb" attributes, indicating its willingness to support
"UDP/TLS/RTP/SAVPF" or "TCP/TLS/RTP/SAVPF". In order to
simplify compatibility with such endpoints, JSEP
implementations MUST follow the following rules when
processing the media m= sections in a received offer:Any profile in the offer matching one of the following
MUST be accepted:
"RTP/AVP" (Defined in
, Section 8.2.2)"RTP/AVPF" (Defined in
, Section 9)"RTP/SAVP" (Defined in
, Section 12)"RTP/SAVPF" (Defined in
, Section 6)"TCP/DTLS/RTP/SAVP" (Defined in
, Section 3.4)"TCP/DTLS/RTP/SAVPF" (Defined in
, Section 3.5)"UDP/TLS/RTP/SAVP" (Defined in
, Section 9)"UDP/TLS/RTP/SAVPF" (Defined in
, Section 9)The profile in any "m=" line in any generated answer
MUST exactly match the profile provided in the offer.Because DTLS-SRTP is REQUIRED, the choice of SAVP or
AVP has no effect; support for DTLS-SRTP is determined by
the presence of one or more "a=fingerprint" attribute.
Note that lack of an "a=fingerprint" attribute will lead
to negotiation failure.The use of AVPF or AVP simply controls the timing
rules used for RTCP feedback. If AVPF is provided, or an
"a=rtcp-fb" attribute is present, assume AVPF timing,
i.e., a default value of "trr-int=0". Otherwise, assume
that AVPF is being used in an AVP compatible mode and use
a value of "trr-int=4000".For data m= sections, implementations MUST support
receiving the "UDP/DTLS/SCTP", "TCP/DTLS/SCTP", or
"DTLS/SCTP" (for backwards compatibility) profiles.Note that re-offers by JSEP implementations MUST use the
correct profile strings even if the initial offer/answer
exchange used an (incorrect) older profile string. This
simplifies JSEP behavior, with minimal downside, as any
remote endpoint that fails to handle such a re-offer will
also fail to handle a JSEP endpoint's initial offer.When createOffer is called, a new SDP description must be
created that includes the functionality specified in
. The exact
details of this process are explained below.When createOffer is called for the first time, the result
is known as the initial offer.The first step in generating an initial offer is to
generate session-level attributes, as specified in
, Section 5. Specifically:
The first SDP line MUST be "v=0", as specified in
, Section 5.1The second SDP line MUST be an "o=" line, as specified
in
, Section 5.2. The value of
the <username> field SHOULD be "-". The sess-id MUST
be representable by a 64-bit signed integer, and the
initial value MUST be less than (2**62)-1, as required by
. It is RECOMMENDED that the
sess-id be constructed by generating a 64-bit quantity with
the two highest bits being set to zero and the remaining 62
bits being cryptographically random. The value of the
<nettype> <addrtype> <unicast-address>
tuple SHOULD be set to a non-meaningful address, such as IN
IP4 0.0.0.0, to prevent leaking the local address in this
field. As mentioned in
, the entire o= line needs to
be unique, but selecting a random number for
<sess-id> is sufficient to accomplish this.The third SDP line MUST be a "s=" line, as specified in
, Section 5.3; to match the
"o=" line, a single dash SHOULD be used as the session
name, e.g. "s=-". Note that this differs from the advice in
which proposes a single space, but
as both "o=" and "s=" are meaningless in JSEP, having the
same meaningless value seems clearer.Session Information ("i="), URI ("u="), Email Address
("e="), Phone Number ("p="), Repeat Times ("r="), and Time
Zones ("z=") lines are not useful in this context and
SHOULD NOT be included.Encryption Keys ("k=") lines do not provide sufficient
security and MUST NOT be included.A "t=" line MUST be added, as specified in
, Section 5.9; both
<start-time> and <stop-time> SHOULD be set to
zero, e.g. "t=0 0".An "a=ice-options" line with the "trickle" option MUST
be added, as specified in
, Section 4.If WebRTC identity is being used, an "a=identity" line
as described in
, Section
5.The next step is to generate m= sections, as specified in
, Section 5.14. An m= section is
generated for each RtpTransceiver that has been added to the
PeerConnection, excluding any stopped RtpTransceivers; this
is done in the order the RtpTransceivers were added to the
PeerConnection. If there are no such RtpTransceivers, no m=
sections are generated; more can be added later, as discussed
in
, Section 5.For each m= section generated for an RtpTransceiver,
establish a mapping between the transceiver and the index of
the generated m= section.Each m= section, provided it is not marked as bundle-only,
MUST generate a unique set of ICE credentials and gather its
own unique set of ICE candidates. Bundle-only m= sections
MUST NOT contain any ICE credentials and MUST NOT gather any
candidates.For DTLS, all m= sections MUST use all the certificate(s)
that have been specified for the PeerConnection; as a result,
they MUST all have the same
fingerprint value(s), or these
value(s) MUST be session-level attributes.Each m= section should be generated as specified in
, Section 5.14. For the m= line
itself, the following rules MUST be followed:
If the m= section is marked as bundle-only, then the
port value MUST be set to 0. Otherwise, the port value is
set to the port of the default ICE candidate for this m=
section, but given that no candidates are available yet,
the "dummy" port value of 9 (Discard) MUST be used, as
indicated in
, Section
5.1.To properly indicate use of DTLS, the <proto>
field MUST be set to "UDP/TLS/RTP/SAVPF", as specified in
, Section 8.If codec preferences have been set for the associated
transceiver, media formats MUST be generated in the
corresponding order, and MUST exclude any codecs not
present in the codec preferences.Unless excluded by the above restrictions, the media
formats MUST include the mandatory audio/video codecs as
specified in
, Section 3, and
, Section 5.The m= line MUST be followed immediately by a "c=" line,
as specified in
, Section 5.7. Again, as no
candidates are available yet, the "c=" line must contain the
"dummy" value "IN IP4 0.0.0.0", as defined in
, Section 5.1. groups
SDP attributes into different categories. To avoid
unnecessary duplication when bundling, attributes of category
IDENTICAL or TRANSPORT MUST NOT be repeated in bundled m=
sections, repeating the guidance from
,
Section 8.1. This includes m= sections for which bundling has
been negotiated and is still desired, as well as m= sections
marked as bundle-only.The following attributes, which are of a category other
than IDENTICAL or TRANSPORT, MUST be included in each m=
section:An "a=mid" line, as specified in
, Section 4. All MID values
MUST be generated in a fashion that does not leak user
information, e.g., randomly or using a per-PeerConnection
counter, and SHOULD be 3 bytes or less, to allow them to
efficiently fit into the RTP header extension defined in
, Section 14. Note that this does not set the
RtpTransceiver mid property, as that only occurs when the
description is applied. The generated MID value can be
considered a "proposed" MID at this point.A direction attribute which is the same as that of the
associated transceiver.For each media format on the m= line, "a=rtpmap" and
"a=fmtp" lines, as specified in
, Section 6, and
, Section 5.1.For each primary codec where RTP retransmission should
be used, a corresponding "a=rtpmap" line indicating "rtx"
with the clock rate of the primary codec and an "a=fmtp"
line that references the payload type of the primary
codec, as specified in
, Section 8.1.For each supported FEC mechanism, "a=rtpmap" and
"a=fmtp" lines, as specified in
, Section 6. The FEC
mechanisms that MUST be supported are specified in
, Section 6,
and specific usage for each media type is outlined in
Sections 4 and 5.If this m= section is for media with configurable
durations of media per packet, e.g., audio, an
"a=maxptime" line, indicating the maximum amount of
media, specified in milliseconds, that can be
encapsulated in each packet, as specified in
, Section 6. This value is
set to the smallest of the maximum duration values across
all the codecs included in the m= section.If this m= section is for video media, and there are
known limitations on the size of images which can be
decoded, an "a=imageattr" line, as specified in
.For each supported RTP header extension, an "a=extmap"
line, as specified in
, Section 5. The list of
header extensions that SHOULD/MUST be supported is
specified in
, Section
5.2. Any header extensions that require encryption MUST
be specified as indicated in
, Section 4.For each supported RTCP feedback mechanism, an
"a=rtcp-fb" line, as specified in
, Section 4.2. The list of
RTCP feedback mechanisms that SHOULD/MUST be supported is
specified in
, Section
5.1.If the RtpTransceiver has a sendrecv or sendonly
direction:
For each MediaStream that was associated with the
transceiver when it was created via addTrack or
addTransceiver, an "a=msid" line, as specified in
, Section 2.
If a MediaStreamTrack is attached to the transceiver's
RtpSender, the "a=msid" lines MUST use that track's ID.
If no MediaStreamTrack is attached, a valid ID MUST be
generated, in the same way that the implementation
generates IDs for local tracks.If no MediaStream is associated with the
transceiver, a single "a=msid" line with the special
value "-" in place of the MediaStream ID, as specified
in
, Section 3.
The track ID MUST be selected as described above.If the RtpTransceiver has a sendrecv or sendonly
direction, and the application has specified RID values
or has specified more than one encoding in the
RtpSenders's parameters, an "a=rid" line for each
encoding specified. The "a=rid" line is specified in
, and its
direction MUST be "send". If the application has chosen a
RID value, it MUST be used as the rid-identifier;
otherwise a RID value MUST be generated by the
implementation. RID values MUST be generated in a fashion
that does not leak user information, e.g., randomly or
using a per-PeerConnection counter, and SHOULD be 3 bytes
or less, to allow them to efficiently fit into the RTP
header extension defined in
, Section 3. If
no encodings have been specified, or only one encoding is
specified but without a RID value, then no "a=rid" lines
are generated.If the RtpTransceiver has a sendrecv or sendonly
direction and more than one "a=rid" line has been
generated, an "a=simulcast" line, with direction "send",
as defined in
,
Section 6.2. The list of RIDs MUST include all of the RID
identifiers used in the "a=rid" lines for this m=
section.If the bundle policy for this PeerConnection is set to
"max-bundle", and this is not the first m= section, or
the bundle policy is set to "balanced", and this is not
the first m= section for this media type, an
"a=bundle-only" line.The following attributes, which are of category IDENTICAL
or TRANSPORT, MUST appear only in "m=" sections which either
have a unique address or which are associated with the
bundle-tag. (In initial offers, this means those "m="
sections which do not contain an "a=bundle-only"
attribute.)"a=ice-ufrag" and "a=ice-pwd" lines, as specified in
, Section 15.4.For each desired digest algorithm, one or more
"a=fingerprint" lines for each of the endpoint's
certificates, as specified in
, Section 5.An "a=setup" line, as specified in
, Section 4, and clarified
for use in DTLS-SRTP scenarios in
, Section 5. The role value
in the offer MUST be "actpass".An "a=tls-id" line, as specified in
, Section
5.2.An "a=rtcp" line, as specified in
, Section 2.1, containing
the dummy value "9 IN IP4 0.0.0.0", because no candidates
have yet been gathered.An "a=rtcp-mux" line, as specified in
, Section 5.1.3.If the RTP/RTCP multiplexing policy is "require", an
"a=rtcp-mux-only" line, as specified in
, Section
4.An "a=rtcp-rsize" line, as specified in
, Section 5.Lastly, if a data channel has been created, a m= section
MUST be generated for data. The <media> field MUST be
set to "application" and the <proto> field MUST be set
to "UDP/DTLS/SCTP"
. The "fmt"
value MUST be set to "webrtc-datachannel" as specified in
, Section
4.1.Within the data m= section, an "a=mid" line MUST be
generated and included as described above, along with an
"a=sctp-port" line referencing the SCTP port number, as
defined in
, Section 5.1,
and, if appropriate, an "a=max-message-size" line, as defined
in
, Section
6.1.As discussed above, the following attributes of category
IDENTICAL or TRANSPORT are included only if the data m=
section either has a unique address or is associated with the
bundle-tag (e.g., if it is the only m= section):
"a=ice-ufrag""a=ice-pwd""a=fingerprint""a=setup""a=tls-id"Once all m= sections have been generated, a session-level
"a=group" attribute MUST be added as specified in
. This attribute MUST have
semantics "BUNDLE", and MUST include the mid identifiers of
each m= section. The effect of this is that the JSEP
implementation offers all m= sections as one bundle group.
However, whether the m= sections are bundle-only or not
depends on the bundle policy.The next step is to generate session-level lip sync groups
as defined in
, Section 7. For each MediaStream
referenced by more than one RtpTransceiver (by passing those
MediaStreams as arguments to the addTrack and addTransceiver
methods), a group of type "LS" MUST be added that contains
the mid values for each RtpTransceiver.Attributes which SDP permits to either be at the session
level or the media level SHOULD generally be at the media
level even if they are identical. This assists development
and debugging by making it easier to understand individual
media sections, especially if one of a set of initially
identical attributes is subsequently changed. However,
implementations MAY choose to aggregate attributes at the
session level and JSEP implementations MUST be prepared to
receive attributes in either location.Attributes other than the ones specified above MAY be
included, except for the following attributes which are
specifically incompatible with the requirements of
, and MUST
NOT be included:
"a=crypto""a=key-mgmt""a=ice-lite"Note that when bundle is used, any additional attributes
that are added MUST follow the advice in
on
how those attributes interact with bundle.Note that these requirements are in some cases stricter
than those of SDP. Implementations MUST be prepared to accept
compliant SDP even if it would not conform to the
requirements for generating SDP in this specification.When createOffer is called a second (or later) time, or is
called after a local description has already been installed,
the processing is somewhat different than for an initial
offer.If the previous offer was not applied using
setLocalDescription, meaning the PeerConnection is still in
the "stable" state, the steps for generating an initial offer
should be followed, subject to the following restriction:
The fields of the "o=" line MUST stay the same except
for the <session-version> field, which MUST increment
by one on each call to createOffer if the offer might
differ from the output of the previous call to createOffer;
implementations MAY opt to increment
<session-version> on every call. The value of the
generated <session-version> is independent of the
<session-version> of the current local description;
in particular, in the case where the current version is N,
an offer is created and applied with version N+1, and then
that offer is rolled back so that the current version is
again N, the next generated offer will still have version
N+2.Note that if the application creates an offer by reading
currentLocalDescription instead of calling createOffer, the
returned SDP may be different than when setLocalDescription
was originally called, due to the addition of gathered ICE
candidates, but the <session-version> will not have
changed. There are no known scenarios in which this causes
problems, but if this is a concern, the solution is simply to
use createOffer to ensure a unique
<session-version>.If the previous offer was applied using
setLocalDescription, but a corresponding answer from the
remote side has not yet been applied, meaning the
PeerConnection is still in the "have-local-offer" state, an
offer is generated by following the steps in the "stable"
state above, along with these exceptions:
The "s=" and "t=" lines MUST stay the same.If any RtpTransceiver has been added, and there exists
an m= section with a zero port in the current local
description or the current remote description, that m=
section MUST be recycled by generating an m= section for
the added RtpTransceiver as if the m= section were being
added to the session description (including a new MID
value), and placing it at the same index as the m= section
with a zero port.If an RtpTransceiver is stopped and is not associated
with an m= section, an m= section MUST NOT be generated for
it. This prevents adding back RtpTransceivers whose m=
sections were recycled and used for a new RtpTransceiver in
a previous offer/ answer exchange, as described above.If an RtpTransceiver has been stopped and is associated
with an m= section, and the m= section is not being
recycled as described above, an m= section MUST be
generated for it with the port set to zero and all "a=msid"
lines removed.For RtpTransceivers that are not stopped, the "a=msid"
line(s) MUST stay the same if they are present in the
current description, regardless of changes to the
transceiver's direction or track. If no "a=msid" line is
present in the current description, "a=msid" line(s) MUST
be generated according to the same rules as for an initial
offer.Each "m=" and c=" line MUST be filled in with the port,
protocol, and address of the default candidate for the m=
section, as described in
, Section 4.3. If ICE
checking has already completed for one or more candidate
pairs and a candidate pair is in active use, then that pair
MUST be used, even if ICE has not yet completed. Note that
this differs from the guidance in
, Section 9.1.2.2, which only
refers to offers created when ICE has completed. In each
case, if no RTP candidates have yet been gathered, dummy
values MUST be used, as described above.Each "a=mid" line MUST stay the same.Each "a=ice-ufrag" and "a=ice-pwd" line MUST stay the
same, unless the ICE configuration has changed (either
changes to the supported STUN/TURN servers, or the ICE
candidate policy), or the "IceRestart" option (
was specified. If the m=
section is bundled into another m= section, it still MUST
NOT contain any ICE credentials.If the m= section is not bundled into another m=
section, its "a=rtcp" attribute line MUST be filled in with
the port and address of the default RTCP candidate, as
indicated in
, Section 5.1.3. If no RTCP
candidates have yet been gathered, dummy values MUST be
used, as described in the initial offer section above.If the m= section is not bundled into another m=
section, for each candidate that has been gathered during
the most recent gathering phase (see
), an
"a=candidate" line MUST be added, as defined in
, Section 4.3., paragraph 3.
If candidate gathering for the section has completed, an
"a=end-of-candidates" attribute MUST be added, as described
in
, Section 9.3.
If the m= section is bundled into another m= section, both
"a=candidate" and "a=end-of-candidates" MUST be
omitted.For RtpTransceivers that are still present, the "a=rid"
lines MUST stay the same.For RtpTransceivers that are still present, any
"a=simulcast" line MUST stay the same.If the previous offer was applied using
setLocalDescription, and a corresponding answer from the
remote side has been applied using setRemoteDescription,
meaning the PeerConnection is in the "have-remote-pranswer"
or "stable" states, an offer is generated based on the
negotiated session descriptions by following the steps
mentioned for the "have-local-offer" state above.In addition, for each existing, non-recycled, non-rejected
m= section in the new offer, the following adjustments are
made based on the contents of the corresponding m= section in
the current local or remote description, as appropriate:
The m= line and corresponding "a=rtpmap" and "a=fmtp"
lines MUST only include media formats which have not been
excluded by the codec preferences of the associated
transceiver, and MUST include all currently available
formats. Media formats that were previously offered but are
no longer available (e.g., a shared hardware codec) MAY be
excluded.Unless codec preferences have been set for the
associated transceiver, the media formats on the m= line
MUST be generated in the same order as in the most recent
answer. Any media formats that were not present in the most
recent answer MUST be added after all existing formats.The RTP header extensions MUST only include those that
are present in the most recent answer.The RTCP feedback mechanisms MUST only include those
that are present in the most recent answer, except for the
case of format-specific mechanisms that are referencing a
newly-added media format.The "a=rtcp" line MUST NOT be added if the most recent
answer included an "a=rtcp-mux" line.The "a=rtcp-mux" line MUST be the same as that in the
most recent answer.The "a=rtcp-mux-only" line MUST NOT be added.The "a=rtcp-rsize" line MUST NOT be added unless present
in the most recent answer.An "a=bundle-only" line MUST NOT be added, as indicated
in
,
Section 6. Instead, JSEP implementations MUST simply omit
parameters in the IDENTICAL and TRANSPORT categories for
bundled m= sections, as described in
,
Section 8.1.Note that if media m= sections are bundled into a data
m= section, then certain TRANSPORT and IDENTICAL attributes
may appear in the data m= section even if they would
otherwise only be appropriate for a media m= section (e.g.,
"a=rtcp-mux"). This cannot happen in initial offers because
in the initial offer JSEP implementations always list media
m= sections (if any) before the data m= section (if any),
and at least one of those media m= sections will not have
the "a=bundle-only" attribute. Therefore, in initial
offers, any "a=bundle-only" m= sections will be bundled
into a preceding non-bundle-only media m= section.The "a=group:BUNDLE" attribute MUST include the MID
identifiers specified in the bundle group in the most recent
answer, minus any m= sections that have been marked as
rejected, plus any newly added or re-enabled m= sections. In
other words, the bundle attribute must contain all m=
sections that were previously bundled, as long as they are
still alive, as well as any new m= sections."a=group:LS" attributes are generated in the same way as
for initial offers, with the additional stipulation that any
lip sync groups that were present in the most recent answer
MUST continue to exist and MUST contain any previously
existing MID identifiers, as long as the identified m=
sections still exist and are not rejected, and the group
still contains at least two MID identifiers. This ensures
that any synchronized "recvonly" m= sections continue to be
synchronized in the new offer.The createOffer method takes as a parameter an
RTCOfferOptions object. Special processing is performed when
generating a SDP description if the following options are
present.If the "IceRestart" option is specified, with a value of
"true", the offer MUST indicate an ICE restart by
generating new ICE ufrag and pwd attributes, as specified
in
, Section 9.1.1.1. If this
option is specified on an initial offer, it has no effect
(since a new ICE ufrag and pwd are already generated).
Similarly, if the ICE configuration has changed, this
option has no effect, since new ufrag and pwd attributes
will be generated automatically. This option is primarily
useful for reestablishing connectivity in cases where
failures are detected by the application.Silence suppression, also known as discontinuous
transmission ("DTX"), can reduce the bandwidth used for
audio by switching to a special encoding when voice
activity is not detected, at the cost of some fidelity.If the "VoiceActivityDetection" option is specified,
with a value of "true", the offer MUST indicate support for
silence suppression in the audio it receives by including
comfort noise ("CN") codecs for each offered audio codec,
as specified in
, Section 5.1, except for
codecs that have their own internal silence suppression
support. For codecs that have their own internal silence
suppression support, the appropriate fmtp parameters for
that codec MUST be specified to indicate that silence
suppression for received audio is desired. For example,
when using the Opus codec
, the "usedtx=1" parameter,
specified in
, would be used in the offer.If the "VoiceActivityDetection" option is specified,
with a value of "false", the JSEP implementation MUST NOT
emit "CN" codecs. For codecs that have their own internal
silence suppression support, the appropriate fmtp
parameters for that codec MUST be specified to indicate
that silence suppression for received audio is not desired.
For example, when using the Opus codec, the "usedtx=0"
parameter would be specified in the offer. In addition, the
implementation MUST NOT use silence suppression for media
it generates, regardless of whether the "CN" codecs or
related fmtp parameters appear in the peer's description.
The impact of these rules is that silence suppression in
JSEP depends on mutual agreement of both sides, which
ensures consistent handling regardless of which codec is
used.The "VoiceActivityDetection" option does not have any
impact on the setting of the "vad" value in the signaling
of the client to mixer audio level header extension
described in
, Section 4.When createAnswer is called, a new SDP description must be
created that is compatible with the supplied remote description
as well as the requirements specified in
. The exact
details of this process are explained below.When createAnswer is called for the first time after a
remote description has been provided, the result is known as
the initial answer. If no remote description has been
installed, an answer cannot be generated, and an error MUST
be returned.Note that the remote description SDP may not have been
created by a JSEP endpoint and may not conform to all the
requirements listed in
. For many cases, this
is not a problem. However, if any mandatory SDP attributes
are missing, or functionality listed as mandatory-to-use
above is not present, this MUST be treated as an error, and
MUST cause the affected m= sections to be marked as
rejected.The first step in generating an initial answer is to
generate session-level attributes. The process here is
identical to that indicated in the initial offers section
above, except that the "a=ice-options" line, with the
"trickle" option as specified in
, Section 4, is
only included if such an option was present in the offer.The next step is to generate session-level lip sync
groups, as defined in
, Section 7. For each group of type
"LS" present in the offer, select the local RtpTransceivers
that are referenced by the MID values in the specified group,
and determine which of them either reference a common local
MediaStream (specified in the calls to
addTrack/addTransceiver used to create them), or have no
MediaStream to reference because they were not created by
addTrack/addTransceiver. If at least two such RtpTransceivers
exist, a group of type "LS" with the mid values of these
RtpTransceivers MUST be added. Otherwise the offered "LS"
group MUST be ignored and no corresponding group generated in
the answer.As a simple example, consider the following offer of a
single audio and single video track contained in the same
MediaStream. SDP lines not relevant to this example have been
removed for clarity. As explained in
, a group of type "LS" has
been added that references each track's RtpTransceiver.If the answerer uses a single MediaStream when it adds its
tracks, both of its transceivers will reference this stream,
and so the subsequent answer will contain a "LS" group
identical to that in the offer, as shown below:However, if the answerer groups its tracks into separate
MediaStreams, its transceivers will reference different
streams, and so the subsequent answer will not contain a "LS"
group.Finally, if the answerer does not add any tracks, its
transceivers will not reference any MediaStreams, causing the
preferences of the offerer to be maintained, and so the
subsequent answer will contain an identical "LS" group.The
example later in this
document shows a more involved case of "LS" group
generation.The next step is to generate m= sections for each m=
section that is present in the remote offer, as specified in
, Section 6. For the purposes
of this discussion, any session-level attributes in the offer
that are also valid as media-level attributes are considered
to be present in each m= section. Each offered m= section
will have an associated RtpTransceiver, as described in
. If there are
more RtpTransceivers than there are m= sections, the
unmatched RtpTransceivers will need to be associated in a
subsequent offer.For each offered m= section, if any of the following
conditions are true, the corresponding m= section in the
answer MUST be marked as rejected by setting the port in the
m= line to zero, as indicated in
, Section 6, and further
processing for this m= section can be skipped:
The associated RtpTransceiver has been stopped.None of the offered media formats are supported and, if
applicable, allowed by codec preferences.The bundle policy is "max-bundle", and this is not the
first m= section or in the same bundle group as the first
m= section.The bundle policy is "balanced", and this is not the
first m= section for this media type or in the same bundle
group as the first m= section for this media type.Otherwise, each m= section in the answer should then be
generated as specified in
, Section 6.1. For the m= line
itself, the following rules must be followed:
The port value would normally be set to the port of the
default ICE candidate for this m= section, but given that
no candidates are available yet, the "dummy" port value of
9 (Discard) MUST be used, as indicated in
, Section
5.1.The <proto> field MUST be set to exactly match the
<proto> field for the corresponding m= line in the
offer.If codec preferences have been set for the associated
transceiver, media formats MUST be generated in the
corresponding order, regardless of what was offered, and
MUST exclude any codecs not present in the codec
preferences.Otherwise, the media formats on the m= line MUST be
generated in the same order as those offered in the current
remote description, excluding any currently unsupported
formats. Any currently available media formats that are not
present in the current remote description MUST be added
after all existing formats.In either case, the media formats in the answer MUST
include at least one format that is present in the offer,
but MAY include formats that are locally supported but not
present in the offer, as mentioned in
, Section 6.1. If no common format
exists, the m= section is rejected as described above.The m= line MUST be followed immediately by a "c=" line,
as specified in
, Section 5.7. Again, as no
candidates are available yet, the "c=" line must contain the
"dummy" value "IN IP4 0.0.0.0", as defined in
, Section 5.1.If the offer supports bundle, all m= sections to be
bundled must use the same ICE credentials and candidates; all
m= sections not being bundled must use unique ICE credentials
and candidates. Each m= section MUST contain the following
attributes (which are of attribute types other than IDENTICAL
and TRANSPORT):
If and only if present in the offer, an "a=mid" line, as
specified in
, Section 9.1. The "mid"
value MUST match that specified in the offer.A direction attribute, determined by applying the rules
regarding the offered direction specified in
, Section 6.1, and then
intersecting with the direction of the associated
RtpTransceiver. For example, in the case where an m=
section is offered as "sendonly", and the local transceiver
is set to "sendrecv", the result in the answer is a
"recvonly" direction.For each media format on the m= line, "a=rtpmap" and
"a=fmtp" lines, as specified in
, Section 6, and
, Section 6.1.If "rtx" is present in the offer, for each primary codec
where RTP retransmission should be used, a corresponding
"a=rtpmap" line indicating "rtx" with the clock rate of the
primary codec and an "a=fmtp" line that references the
payload type of the primary codec, as specified in
, Section 8.1.For each supported FEC mechanism, "a=rtpmap" and
"a=fmtp" lines, as specified in
, Section 6. The FEC
mechanisms that MUST be supported are specified in
, Section 6, and
specific usage for each media type is outlined in Sections
4 and 5.If this m= section is for media with configurable
durations of media per packet, e.g., audio, an "a=maxptime"
line, as described in
.If this m= section is for video media, and there are
known limitations on the size of images which can be
decoded, an "a=imageattr" line, as specified in
.For each supported RTP header extension that is present
in the offer, an "a=extmap" line, as specified in
, Section 5. The list of
header extensions that SHOULD/MUST be supported is
specified in
, Section
5.2. Any header extensions that require encryption MUST be
specified as indicated in
, Section 4.For each supported RTCP feedback mechanism that is
present in the offer, an "a=rtcp-fb" line, as specified in
, Section 4.2. The list of
RTCP feedback mechanisms that SHOULD/MUST be supported is
specified in
, Section
5.1.If the RtpTransceiver has a sendrecv or sendonly
direction:
For each MediaStream that was associated with the
transceiver when it was created via addTrack or
addTransceiver, an "a=msid" line, as specified in
, Section 2.
If a MediaStreamTrack is attached to the transceiver's
RtpSender, the "a=msid" lines MUST use that track's ID.
If no MediaStreamTrack is attached, a valid ID MUST be
generated, in the same way that the implementation
generates IDs for local tracks.If no MediaStream is associated with the transceiver,
a single "a=msid" line with the special value "-" in
place of the MediaStream ID, as specified in
, Section 3.
The track ID MUST be selected as described above.Each m= section which is not bundled into another m=
section, MUST contain the following attributes (which are of
category IDENTICAL or TRANSPORT):"a=ice-ufrag" and "a=ice-pwd" lines, as specified in
, Section 15.4.For each desired digest algorithm, one or more
"a=fingerprint" lines for each of the endpoint's
certificates, as specified in
, Section 5.An "a=setup" line, as specified in
, Section 4, and clarified
for use in DTLS-SRTP scenarios in
, Section 5. The role value
in the answer MUST be "active" or "passive". When the
offer contains the "actpass" value, as will always be the
case with JSEP endpoints, the answerer SHOULD use the
"active" role. Offers from non-JSEP endpoints MAY send
other values for "a=setup", in which case the answer MUST
use a value consistent with the value in the offer.An "a=tls-id" line, as specified in
, Section
5.3.If present in the offer, an "a=rtcp-mux" line, as
specified in
, Section 5.1.3. Otherwise,
an "a=rtcp" line, as specified in
, Section 2.1, containing
the dummy value "9 IN IP4 0.0.0.0" (because no candidates
have yet been gathered).If present in the offer, an "a=rtcp-rsize" line, as
specified in
, Section 5.If a data channel m= section has been offered, a m=
section MUST also be generated for data. The <media>
field MUST be set to "application" and the <proto> and
<fmt> fields MUST be set to exactly match the fields in
the offer.Within the data m= section, an "a=mid" line MUST be
generated and included as described above, along with an
"a=sctp-port" line referencing the SCTP port number, as
defined in
, Section 5.1,
and, if appropriate, an "a=max-message-size" line, as defined
in
, Section
6.1.As discussed above, the following attributes of category
IDENTICAL or TRANSPORT are included only if the data m=
section is not bundled into another m= section:
"a=ice-ufrag""a=ice-pwd""a=fingerprint""a=setup""a=tls-id"Note that if media m= sections are bundled into a data m=
section, then certain TRANSPORT and IDENTICAL attributes may
also appear in the data m= section even if they would
otherwise only be appropriate for a media m= section (e.g.,
"a=rtcp-mux").If "a=group" attributes with semantics of "BUNDLE" are
offered, corresponding session-level "a=group" attributes
MUST be added as specified in
. These attributes MUST have
semantics "BUNDLE", and MUST include the all mid identifiers
from the offered bundle groups that have not been rejected.
Note that regardless of the presence of "a=bundle-only" in
the offer, no m= sections in the answer should have an
"a=bundle-only" line.Attributes that are common between all m= sections MAY be
moved to session-level, if explicitly defined to be valid at
session-level.The attributes prohibited in the creation of offers are
also prohibited in the creation of answers.When createAnswer is called a second (or later) time, or
is called after a local description has already been
installed, the processing is somewhat different than for an
initial answer.If the previous answer was not applied using
setLocalDescription, meaning the PeerConnection is still in
the "have-remote-offer" state, the steps for generating an
initial answer should be followed, subject to the following
restriction:
The fields of the "o=" line MUST stay the same except
for the <session-version> field, which MUST increment
if the session description changes in any way from the
previously generated answer.If any session description was previously supplied to
setLocalDescription, an answer is generated by following the
steps in the "have-remote-offer" state above, along with
these exceptions:
The "s=" and "t=" lines MUST stay the same.Each "m=" and c=" line MUST be filled in with the port
and address of the default candidate for the m= section, as
described in
, Section 4.3. Note, however,
that the m= line protocol need not match the default
candidate, because this protocol value must instead match
what was supplied in the offer, as described above.Each "a=ice-ufrag" and "a=ice-pwd" line MUST stay the
same, unless the m= section is restarting, in which case
new ICE credentials must be created as specified in
, Section 9.2.1.1. If the m=
section is bundled into another m= section, it still MUST
NOT contain any ICE credentials.Each "a=tls-id" line MUST stay the same unless the
offerer's "a=tls-id" line changed, in which case a new
"a=tls-id" value MUST be created, as described in
, Section
5.2.Each "a=setup" line MUST use an "active" or "passive"
role value consistent with the existing DTLS association,
if the association is being continued by the offerer.RTCP multiplexing must be used, and an "a=rtcp-mux" line
inserted if and only if the m= section previously used RTCP
multiplexing.If the m= section is not bundled into another m= section
and RTCP multiplexing is not active, an "a=rtcp" attribute
line MUST be filled in with the port and address of the
default RTCP candidate. If no RTCP candidates have yet been
gathered, dummy values MUST be used, as described in the
initial answer section above.If the m= section is not bundled into another m=
section, for each candidate that has been gathered during
the most recent gathering phase (see
), an
"a=candidate" line MUST be added, as defined in
, Section 4.3., paragraph 3.
If candidate gathering for the section has completed, an
"a=end-of-candidates" attribute MUST be added, as described
in
, Section 9.3.
If the m= section is bundled into another m= section, both
"a=candidate" and "a=end-of-candidates" MUST be
omitted.For RtpTransceivers that are not stopped, the "a=msid"
line(s) MUST stay the same, regardless of changes to the
transceiver's direction or track. If no "a=msid" line is
present in the current description, "a=msid" line(s) MUST
be generated according to the same rules as for an initial
answer.The createAnswer method takes as a parameter an
RTCAnswerOptions object. The set of parameters for
RTCAnswerOptions is different than those supported in
RTCOfferOptions; the IceRestart option is unnecessary, as ICE
credentials will automatically be changed for all m= sections
where the offerer chose to perform ICE restart.The following options are supported in
RTCAnswerOptions.Silence suppression in the answer is handled as
described in
, with
one exception: if support for silence suppression was not
indicated in the offer, the VoiceActivityDetection
parameter has no effect, and the answer should be generated
as if VoiceActivityDetection was set to false. This is done
on a per-codec basis (e.g., if the offerer somehow offered
support for CN but set "usedtx=0" for Opus, setting
VoiceActivityDetection to true would result in an answer
with CN codecs and "usedtx=0"). The impact of this rule is
that an answerer will not try to use silence suppression
with any endpoint that does not offer it, making silence
suppression support bilateral even with non-JSEP
endpoints.The SDP returned from createOffer or createAnswer MUST NOT
be changed before passing it to setLocalDescription. If precise
control over the SDP is needed, the aforementioned
createOffer/createAnswer options or RtpTransceiver APIs MUST be
used.After calling setLocalDescription with an offer or answer,
the application MAY modify the SDP to reduce its capabilities
before sending it to the far side, as long as it follows the
rules above that define a valid JSEP offer or answer. Likewise,
an application that has received an offer or answer from a peer
MAY modify the received SDP, subject to the same constraints,
before calling setRemoteDescription.As always, the application is solely responsible for what it
sends to the other party, and all incoming SDP will be
processed by the JSEP implementation to the extent of its
capabilities. It is an error to assume that all SDP is
well-formed; however, one should be able to assume that any
implementation of this specification will be able to process,
as a remote offer or answer, unmodified SDP coming from any
other implementation of this specification.When a SessionDescription is supplied to
setLocalDescription, the following steps MUST be performed:
If the description is of type "rollback", follow the
processing defined in
and skip the
processing described in the rest of this section.Otherwise, the type of the SessionDescription is checked
against the current state of the PeerConnection:
If the type is "offer", the PeerConnection state MUST be
either "stable" or "have-local-offer".If the type is "pranswer" or "answer", the
PeerConnection state MUST be either "have-remote-offer" or
"have-local-pranswer".If the type is not correct for the current state,
processing MUST stop and an error MUST be returned.The SessionDescription is then checked to ensure that its
contents are identical to those generated in the last call to
createOffer/createAnswer, and thus have not been altered, as
discussed in
; otherwise, processing
MUST stop and an error MUST be returned.Next, the SessionDescription is parsed into a data
structure, as described in
below.Finally, the parsed SessionDescription is applied as
described in
below.When a SessionDescription is supplied to
setRemoteDescription, the following steps MUST be performed:
If the description is of type "rollback", follow the
processing defined in
and skip the
processing described in the rest of this section.Otherwise, the type of the SessionDescription is checked
against the current state of the PeerConnection:
If the type is "offer", the PeerConnection state MUST be
either "stable" or "have-remote-offer".If the type is "pranswer" or "answer", the
PeerConnection state MUST be either "have-local-offer" or
"have-remote-pranswer".If the type is not correct for the current state,
processing MUST stop and an error MUST be returned.Next, the SessionDescription is parsed into a data
structure, as described in
below. If parsing fails
for any reason, processing MUST stop and an error MUST be
returned.Finally, the parsed SessionDescription is applied as
described in
below.A rollback may be performed if the PeerConnection is in any
state except for "stable". This means that both offers and
provisional answers can be rolled back. Rollback can only be
used to cancel proposed changes; there is no support for
rolling back from a stable state to a previous stable state. If
a rollback is attempted in the "stable" state, processing MUST
stop and an error MUST be returned. Note that this implies that
once the answerer has performed setLocalDescription with his
answer, this cannot be rolled back.The effect of rollback MUST be the same regardless of
whether setLocalDescription or setRemoteDescription is
called.In order to process rollback, a JSEP implementation abandons
the current offer/answer transaction, sets the signaling state
to "stable", and sets the pending local and/or remote
description (see
and
) to null. Any
resources or candidates that were allocated by the abandoned
local description are discarded; any media that is received is
processed according to the previous local and remote
descriptions.A rollback disassociates any RtpTransceivers that were
associated with m= sections by the application of the
rolled-back session description (see
and
). This means that
some RtpTransceivers that were previously associated will no
longer be associated with any m= section; in such cases, the
value of the RtpTransceiver's mid property MUST be set to null,
and the mapping between the transceiver and its m= section
index MUST be discarded. RtpTransceivers that were created by
applying a remote offer that was subsequently rolled back MUST
be stopped and removed from the PeerConnection. However, a
RtpTransceiver MUST NOT be removed if a track was attached to
the RtpTransceiver via the addTrack method. This is so that an
application may call addTrack, then call setRemoteDescription
with an offer, then roll back that offer, then call createOffer
and have a m= section for the added track appear in the
generated offer.The SDP contained in the session description object consists
of a sequence of text lines, each containing a key-value
expression, as described in
, Section 5. The SDP is read,
line-by-line, and converted to a data structure that contains
the deserialized information. However, SDP allows many types of
lines, not all of which are relevant to JSEP applications. For
each line, the implementation will first ensure it is
syntactically correct according to its defining ABNF, check
that it conforms to
and
semantics, and then either parse and
store or discard the provided value, as described below.If any line is not well-formed, or cannot be parsed as
described, the parser MUST stop with an error and reject the
session description, even if the value is to be discarded. This
ensures that implementations do not accidentally misinterpret
ambiguous SDP.First, the session-level lines are checked and parsed.
These lines MUST occur in a specific order, and with a
specific syntax, as defined in
, Section 5. Note that while the
specific line types (e.g. "v=", "c=") MUST occur in the
defined order, lines of the same type (typically "a=") can
occur in any order.The following non-attribute lines are not meaningful in
the JSEP context and MAY be discarded once they have been
checked.
The "c=" line MUST be checked for syntax but its value
is only used for ICE mismatch detection, as defined in
, Section 6.1. Note that JSEP
implementations should never encounter this condition
because ICE is required for WebRTC.The "i=", "u=", "e=", "p=", "t=", "r=", "z=", and "k="
lines are not used by this specification; they MUST be
checked for syntax but their values are not used.The remaining non-attribute lines are processed as
follows:
The "v=" line MUST have a version of 0, as specified in
, Section 5.1.The "o=" line MUST be parsed as specified in
, Section 5.2.The "b=" line, if present, MUST be parsed as specified
in
, Section 5.8, and the bwtype and
bandwidth values stored.Finally, the attribute lines are processed. Specific
processing MUST be applied for the following session-level
attribute ("a=") lines:
Any "a=group" lines are parsed as specified in
, Section 5, and the group's
semantics and mids are stored.If present, a single "a=ice-lite" line is parsed as
specified in
, Section 15.3, and a value
indicating the presence of ice-lite is stored.If present, a single "a=ice-ufrag" line is parsed as
specified in
, Section 15.4, and the ufrag
value is stored.If present, a single "a=ice-pwd" line is parsed as
specified in
, Section 15.4, and the password
value is stored.If present, a single "a=ice-options" line is parsed as
specified in
, Section 15.5, and the set of
specified options is stored.Any "a=fingerprint" lines are parsed as specified in
, Section 5, and the set of
fingerprint and algorithm values is stored.If present, a single "a=setup" line is parsed as
specified in
, Section 4, and the setup value
is stored.If present, a single "a=tls-id" line is parsed as
specified in
Section 5, and
the tls-id value is stored.Any "a=identity" lines are parsed and the identity
values stored for subsequent verification, as specified
, Section
5.Any "a=extmap" lines are parsed as specified in
, Section 5, and their values are
stored.Other attributes that are not relevant to JSEP may also be
present, and implementations SHOULD process any that they
recognize. As required by
, Section 5.13, unknown
attribute lines MUST be ignored.Once all the session-level lines have been parsed,
processing continues with the lines in m= sections.Like the session-level lines, the media section lines MUST
occur in the specific order and with the specific syntax
defined in
, Section 5.The "m=" line itself MUST be parsed as described in
, Section 5.14, and the media, port,
proto, and fmt values stored.Following the "m=" line, specific processing MUST be
applied for the following non-attribute lines:
As with the "c=" line at the session level, the "c="
line MUST be parsed according to
, Section 5.7, but its value is
not used.The "b=" line, if present, MUST be parsed as specified
in
, Section 5.8, and the bwtype and
bandwidth values stored.Specific processing MUST also be applied for the following
attribute lines:
If present, a single "a=ice-ufrag" line is parsed as
specified in
, Section 15.4, and the ufrag
value is stored.If present, a single "a=ice-pwd" line is parsed as
specified in
, Section 15.4, and the password
value is stored.If present, a single "a=ice-options" line is parsed as
specified in
, Section 15.5, and the set of
specified options is stored.Any "a=candidate" attributes MUST be parsed as specified
in
, Section 15.1, and their values
stored.Any "a=remote-candidates" attributes MUST be parsed as
specified in
, Section 15.2, but their values
are ignored.If present, a single "a=end-of-candidates" attribute
MUST be parsed as specified in
, Section 8.2, and
its presence or absence flagged and stored.Any "a=fingerprint" lines are parsed as specified in
, Section 5, and the set of
fingerprint and algorithm values is stored.If the "m=" proto value indicates use of RTP, as described
in
above, the following
attribute lines MUST be processed:
The "m=" fmt value MUST be parsed as specified in
, Section 5.14, and the individual
values stored.Any "a=rtpmap" or "a=fmtp" lines MUST be parsed as
specified in
, Section 6, and their values
stored.If present, a single "a=ptime" line MUST be parsed as
described in
, Section 6, and its value
stored.If present, a single "a=maxptime" line MUST be parsed as
described in
, Section 6, and its value
stored.If present, a single direction attribute line (e.g.
"a=sendrecv") MUST be parsed as described in
, Section 6, and its value
stored.Any "a=ssrc" attributes MUST be parsed as specified in
, Section 4.1, and their values
stored.Any "a=extmap" attributes MUST be parsed as specified in
, Section 5, and their values
stored.Any "a=rtcp-fb" attributes MUST be parsed as specified
in
, Section 4.2., and their values
stored.If present, a single "a=rtcp-mux" attribute MUST be
parsed as specified in
, Section 5.1.3, and its
presence or absence flagged and stored.If present, a single "a=rtcp-mux-only" attribute MUST be
parsed as specified in
, Section 3,
and its presence or absence flagged and stored.If present, a single "a=rtcp-rsize" attribute MUST be
parsed as specified in
, Section 5, and its presence or
absence flagged and stored.If present, a single "a=rtcp" attribute MUST be parsed
as specified in
, Section 2.1, but its value is
ignored, as this information is superfluous when using
ICE.If present, "a=msid" attributes MUST be parsed as
specified in
, Section 3.2, and
their values stored.Any "a=imageattr" attributes MUST be parsed as specified
in
, Section 3, and their values
stored.Any "a=rid" lines MUST be parsed as specified in
, Section 10, and
their values stored.If present, a single "a=simulcast" line MUST be parsed
as specified in
, and
its values stored.Otherwise, if the "m=" proto value indicates use of SCTP,
the following attribute lines MUST be processed:
The "m=" fmt value MUST be parsed as specified in
, Section 4.3,
and the application protocol value stored.An "a=sctp-port" attribute MUST be present, and it MUST
be parsed as specified in
, Section 5.2,
and the value stored.If present, a single "a=max-message-size" attribute MUST
be parsed as specified in
, Section 6, and
the value stored. Otherwise, use the specified default.Other attributes that are not relevant to JSEP may also be
present, and implementations SHOULD process any that they
recognize. As required by
, Section 5.13, unknown
attribute lines MUST be ignored.Assuming parsing completes successfully, the parsed
description is then evaluated to ensure internal consistency
as well as proper support for mandatory features.
Specifically, the following checks are performed:
For each m= section, valid values for each of the
mandatory-to-use features enumerated in
MUST be present.
These values MAY either be present at the media level, or
inherited from the session level.
ICE ufrag and password values, which MUST comply with
the size limits specified in
, Section 15.4.tls-id value, which MUST be set according to
, Section 5. If
this is a re-offer or a response to a re-offer and the
tls-id value is different from that presently in use, the
DTLS connection is not being continued and the remote
description MUST be part of an ICE restart, together with
new ufrag and password values.DTLS setup value, which MUST be set according to the
rules specified in
, Section 5 and MUST be
consistent with the selected role of the current DTLS
connection, if one exists and is being continued.DTLS fingerprint values, where at least one
fingerprint MUST be present.All RID values referenced in an "a=simulcast" line MUST
exist as "a=rid" lines.Each m= section is also checked to ensure prohibited
features are not used.If the RTP/RTCP multiplexing policy is "require", each
m= section MUST contain an "a=rtcp-mux" attribute. If an m=
section contains an "a=rtcp-mux-only" attribute, that
section MUST also contain an "a=rtcp-mux" attribute.If an m= section was present in the previous answer, the
state of RTP/RTCP multiplexing MUST match what was
previously negotiated.If this session description is of type "pranswer" or
"answer", the following additional checks are applied:
The session description must follow the rules defined in
, Section 6, including the
requirement that the number of m= sections MUST exactly
match the number of m= sections in the associated
offer.For each m= section, the media type and protocol values
MUST exactly match the media type and protocol values in
the corresponding m= section in the associated offer.If any of the preceding checks failed, processing MUST
stop and an error MUST be returned.The following steps are performed at the media engine level
to apply a local description. If an error is returned, the
session MUST be restored to the state it was in before
performing these steps.First, m= sections are processed. For each m= section, the
following steps MUST be performed; if any parameters are out of
bounds, or cannot be applied, processing MUST stop and an error
MUST be returned.
If this m= section is new, begin gathering candidates for
it, as defined in
, Section 4.1.1, unless it is
definitively being bundled (either this is an offer and the
m= section is marked bundle-only, or it is an answer and the
m= section is bundled into into another m= section.)Or, if the ICE ufrag and password values have changed,
trigger the ICE agent to start an ICE restart, and begin
gathering new candidates for the m= section as described in
, Section 9.1.1.1. If this
description is an answer, also start checks on that media
section as defined in
, Section 9.3.1.1.If the m= section proto value indicates use of RTP:
If there is no RtpTransceiver associated with this m=
section, find one and associate it with this m= section
according to the following steps. Note that this situation
will only occur when applying an offer.
Find the RtpTransceiver that corresponds to this m=
section, using the mapping between transceivers and m=
section indices established when creating the offer.Set the value of this RtpTransceiver's mid property to
the MID of the m= section.If RTCP mux is indicated, prepare to demux RTP and RTCP
from the RTP ICE component, as specified in
, Section 5.1.3.For each specified RTP header extension, establish a
mapping between the extension ID and URI, as described in
, Section 6.If the MID header extension is supported, prepare to
demux RTP streams intended for this m= section based on the
MID header extension, as described in
,
Section 15.For each specified media format, establish a mapping
between the payload type and the actual media format, as
described in
, Section 6.1. In addition,
prepare to demux RTP streams intended for this m= section
based on the media formats supported by this m= section, as
described in
,
Section 10.2.For each specified "rtx" media format, establish a
mapping between the RTX payload type and its associated
primary payload type, as described in
, Sections 8.6 and 8.7.If the directional attribute is of type "sendrecv" or
"recvonly", enable receipt and decoding of media.Finally, if this description is of type "pranswer" or
"answer", follow the processing defined in
below.The following steps are performed to apply a remote
description. If an error is returned, the session MUST be
restored to the state it was in before performing these
steps.If the answer contains any "a=ice-options" attributes where
"trickle" is listed as an attribute, update the PeerConnection
canTrickle property to be true. Otherwise, set this property to
false.The following steps MUST be performed for attributes at the
session level; if any parameters are out of bounds, or cannot
be applied, processing MUST stop and an error MUST be returned.
For any specified "CT" bandwidth value, set this as the
limit for the maximum total bitrate for all m= sections, as
specified in
, Section 5.8. Within this
overall limit, the implementation can dynamically decide how
to best allocate the available bandwidth between m= sections,
respecting any specific limits that have been specified for
individual m= sections.For any specified "RR" or "RS" bandwidth values, handle as
specified in
, Section 2.Any "AS" bandwidth value MUST be ignored, as the meaning
of this construct at the session level is not well
defined.For each m= section, the following steps MUST be performed;
if any parameters are out of bounds, or cannot be applied,
processing MUST stop and an error MUST be returned.
If the ICE ufrag or password changed from the previous
remote description:
.
If the description is of type "offer", the
implementation MUST note that an ICE restart is needed, as
described in
, Section 9.1.1.1.If the description is of type "answer" or "pranswer",
then check to see if the current local description is an
ICE restart, and if not, generate an error. If the
PeerConnection state is "have-remote-pranswer", and the ICE
ufrag or password changed from the previous provisional
answer, then signal the ICE agent to discard any previous
ICE check list state for the m= section. Finally, signal
the ICE agent to begin checks as described in
, Section 9.3.1.1.If the current local description indicates an ICE restart,
and either the ICE ufrag or password has not changed from the
previous remote description, as prescribed by
, Section 9.2.1.1, generate an
error.Configure the ICE components associated with this media
section to use the supplied ICE remote ufrag and password for
their connectivity checks.Pair any supplied ICE candidates with any gathered local
candidates, as described in
, Section 5.7, and start
connectivity checks with the appropriate credentials.If an "a=end-of-candidates" attribute is present, process
the end-of-candidates indication as described in
, Section 11.If the m= section proto value indicates use of RTP:
If the m= section is being recycled (see
), dissociate
the currently associated RtpTransceiver by setting its mid
property to null, and discard the mapping between the
transceiver and its m= section index.If the m= section is not associated with any
RtpTransceiver (possibly because it was dissociated in the
previous step), either find an RtpTransceiver or create one
according to the following steps:
If the m= section is sendrecv or recvonly, and there
are RtpTransceivers of the same type that were added to
the PeerConnection by addTrack and are not associated
with any m= section and are not stopped, find the first
(according to the canonical order described in
) such
RtpTransceiver.If no RtpTransceiver was found in the previous step,
create one with a recvonly direction.Associate the found or created RtpTransceiver with the
m= section by setting the value of the RtpTransceiver's
mid property to the MID of the m= section, and establish
a mapping between the transceiver and the index of the m=
section. If the m= section does not include a MID (i.e.,
the remote endpoint does not support the MID extension),
generate a value for the RtpTransceiver mid property,
following the guidance for "a=mid" mentioned in
.For each specified media format that is also supported
by the local implementation, establish a mapping between
the specified payload type and the media format, as
described in
, Section 6.1. Specifically, this
means that the implementation records the payload type to
be used in outgoing RTP packets when sending each specified
media format, as well as the relative preference for each
format that is indicated in their ordering. If any
indicated media format is not supported by the local
implementation, it MUST be ignored.For each specified "rtx" media format, establish a
mapping between the RTX payload type and its associated
primary payload type, as described in
, Section 4. If any referenced
primary payload types are not present, this MUST result in
an error. Note that RTX payload types may refer to primary
payload types which are not supported by the local media
implementation, in which case, the RTX payload type MUST
also be ignored.For each specified fmtp parameter that is supported by
the local implementation, enable them on the associated
media formats.For each specified SSRC that is signaled in the m=
section, prepare to demux RTP streams intended for this m=
section using that SSRC, as described in
,
Section 10.2.For each specified RTP header extension that is also
supported by the local implementation, establish a mapping
between the extension ID and URI, as described in
, Section 5. Specifically, this
means that the implementation records the extension ID to
be used in outgoing RTP packets when sending each specified
header extension. If any indicated RTP header extension is
not supported by the local implementation, it MUST be
ignored.For each specified RTCP feedback mechanism that is
supported by the local implementation, enable them on the
associated media formats.For any specified "TIAS" bandwidth value, set this value
as a constraint on the maximum RTP bitrate to be used when
sending media, as specified in
. If a "TIAS" value is not
present, but an "AS" value is specified, generate a "TIAS"
value using this formula:
TIAS = AS * 1000 * 0.95 - (50 * 40 * 8)The 50 is based on 50 packets per second, the 40 is
based on an estimate of total header size, the 1000 changes
the unit from kbps to bps (as required by TIAS), and the
0.95 is to allocate 5% to RTCP. "TIAS" is used in
preference to "AS" because it provides more accurate
control of bandwidth.For any "RR" or "RS" bandwidth values, handle as
specified in
, Section 2.Any specified "CT" bandwidth value MUST be ignored, as
the meaning of this construct at the media level is not
well defined.If the m= section is of type audio:
For each specified "CN" media format, configure
silence suppression for all supported media formats with
the same clockrate, as described in
, Section 5, except for formats
that have their own internal silence suppression
mechanisms. Silence suppression for such formats (e.g.,
Opus) is controlled via fmtp parameters, as discussed in
.For each specified "telephone-event" media format,
enable DTMF transmission for all supported media formats
with the same clockrate, as described in
, Section 2.5.1.2. If there are
any supported media formats that do not have a
corresponding telephone-event format, disable DTMF
transmission for those formats.For any specified "ptime" value, configure the
available media formats to use the specified packet size
when sending. If the specified size is not supported for
a media format, use the next closest value instead.Finally, if this description is of type "pranswer" or
"answer", follow the processing defined in
below.In addition to the steps mentioned above for processing a
local or remote description, the following steps are performed
when processing a description of type "pranswer" or
"answer".For each m= section, the following steps MUST be performed:
If the m= section has been rejected (i.e. port is set to
zero in the answer), stop any reception or transmission of
media for this section, and, unless a non-rejected m= section
is bundled with this m= section, discard any associated ICE
components, as described in
, Section 9.2.1.3.If the remote DTLS fingerprint has been changed or the
tls-id has changed, tear down the DTLS connection. This
includes the case when the PeerConnection state is
"have-remote-pranswer". If a DTLS connection needs to be torn
down but the answer does not indicate an ICE restart or, in
the case of "have-remote-pranswer", new ICE credentials, an
error MUST be generated. If an ICE restart is performed
without a change in tls-id or fingerprint, then the same DTLS
connection is continued over the new ICE channel. Note that
although JSEP requires that answerers change the tls-id value
if and only if the offerer does, non-JSEP answerers are
permitted to change the tls-id as long as the offer contained
an ICE restart. Thus, JSEP implementations which process DTLS
data prior to receiving an answer MUST be prepared to receive
either a ClientHello or data from the previous DTLS
connection.If no valid DTLS connection exists, prepare to start a
DTLS connection, using the specified roles and fingerprints,
on any underlying ICE components, once they are active.If the m= section proto value indicates use of RTP:
If the m= section references RTCP feedback mechanisms
that were not present in the corresponding m= section in
the offer, this indicates a negotiation problem and MUST
result in an error. However, new media formats and new RTP
header extension values are permitted in the answer, as
described in
, Section 7, and
, Section 6.If the m= section has RTCP mux enabled, discard the RTCP
ICE component, if one exists, and begin or continue muxing
RTCP over the RTP ICE component, as specified in
, Section 5.1.3. Otherwise,
prepare to transmit RTCP over the RTCP ICE component; if no
RTCP ICE component exists, because RTCP mux was previously
enabled, this MUST result in an error.If the m= section has reduced-size RTCP enabled,
configure the RTCP transmission for this m= section to use
reduced-size RTCP, as specified in
.If the directional attribute in the answer indicates
that the JSEP implementation should be sending media
("sendonly" for local answers, "recvonly" for remote
answers, or "sendrecv" for either type of answer), choose
the media format to send as the most preferred media format
from the remote description that is also locally supported,
as discussed in
, Sections 6.1 and 7, and start
transmitting RTP media using that format once the
underlying transport layers have been established. If an
SSRC has not already been chosen for this outgoing RTP
stream, choose a random one. If media is already being
transmitted, the same SSRC SHOULD be used unless the
clockrate of the new codec is different, in which case a
new SSRC MUST be chosen, as specified in
, Section 3.1.The payload type mapping from the remote description is
used to determine payload types for the outgoing RTP
streams, including the payload type for the send media
format chosen above. Any RTP header extensions that were
negotiated should be included in the outgoing RTP streams,
using the extension mapping from the remote description; if
the RID header extension has been negotiated, and RID
values are specified, include the RID header extension in
the outgoing RTP streams, as indicated in
, Section 4.If the m= section is of type audio, and silence
suppression was configured for the send media format as a
result of processing the remote description, and is also
enabled for that format in the local description, use
silence suppression for outgoing media, in accordance with
the guidance in
. If these
conditions are not met, silence suppression MUST NOT be
used for outgoing media.If simulcast has been negotiated, send the number of
Source RTP Streams as specified in
,
Section 6.2.2.If the send media format chosen above has a
corresponding "rtx" media format, or a FEC mechanism has
been negotiated, establish a Redundancy RTP Stream with a
random SSRC for each Source RTP Stream, and start or
continue transmitting RTX/FEC packets as needed.If the send media format chosen above has a
corresponding "red" media format of the same clockrate,
allow redundant encoding using the specified format for
resiliency purposes, as discussed in
, Section 3.2. Note
that unlike RTX or FEC media formats, the "red" format is
transmitted on the Source RTP Stream, not the Redundancy
RTP Stream.Enable the RTCP feedback mechanisms referenced in the
media section for all Source RTP Streams using the
specified media formats. Specifically, begin or continue
sending the requested feedback types and reacting to
received feedback, as specified in
, Section 4.2. When sending RTCP
feedback, follow the rules and recommendations from
Section 5.4.1, to select
which SSRC to use.If the directional attribute in the answer indicates
that the JSEP implementation should not be sending media
("recvonly" for local answers, "sendonly" for remote
answers, or "inactive" for either type of answer) stop
transmitting all RTP media, but continue sending RTCP, as
described in
, Section 5.1.If the m= section proto value indicates use of SCTP:
If an SCTP association exists, and the remote SCTP port
has changed, discard the existing SCTP association. This
includes the case when the PeerConnection state is
"have-remote-pranswer".If no valid SCTP association exists, prepare to initiate
a SCTP association over the associated ICE component and
DTLS connection, using the local SCTP port value from the
local description, and the remote SCTP port value from the
remote description, as described in
, Section
10.2.If the answer contains valid bundle groups, discard any ICE
components for the m= sections that will be bundled onto the
primary ICE components in each bundle, and begin muxing these
m= sections accordingly, as described in
,
Section 8.2.If the description is of type "answer", and there are still
remaining candidates in the ICE candidate pool, discard
them.When bundling, associating incoming RTP/RTCP with the proper
m= section is defined in
, Section
10.2. When not bundling, the proper m= section is clear from the
ICE component over which the RTP/RTCP is received.Once the proper m= section(s) are known, RTP/RTCP is delivered
to the RtpTransceiver(s) associated with the m= section(s) and
further processing of the RTP/RTCP is done at the RtpTransceiver
level. This includes using RID
to distinguish between
multiple Encoded Streams, as well as determine which Source RTP
stream should be repaired by a given Redundancy RTP stream.Note that this example section shows several SDP fragments. To
format in 72 columns, some of the lines in SDP have been split
into multiple lines, where leading whitespace indicates that a
line is a continuation of the previous line. In addition, some
blank lines have been added to improve readability but are not
valid in SDP.More examples of SDP for WebRTC call flows, including examples
with IPv6 addresses, can be found in
.This section shows a very simple example that sets up a
minimal audio / video call between two JSEP endpoints without
using trickle ICE. The example in the following section
provides a more detailed example of what could happen in a JSEP
session.The code flow below shows Alice's endpoint initiating the
session to Bob's endpoint. The messages from the JavaScript
application in Alice's browser to the JavaScript in Bob's
browser, abbreviated as AliceJS and BobJS respectively, are
assumed to flow over some signaling protocol via a web server.
The JavaScript on both Alice's side and Bob's side waits for
all candidates before sending the offer or answer, so the
offers and answers are complete; trickle ICE is not used. The
user agents (JSEP implementations) in Alice and Bob's browsers,
abbreviated as AliceUA and BobUA respectively, are using the
default bundle policy of "balanced", and the default RTCP mux
policy of "require".The SDP for |offer-A1| looks like:The SDP for |answer-A1| looks like:This section shows a more involved example of a session
between two JSEP endpoints. Trickle ICE is used in full trickle
mode, with a bundle policy of "max-bundle", an RTCP mux policy
of "require", and a single TURN server. Initially, both Alice
and Bob establish an audio channel and a data channel. Later,
Bob adds two video flows, one for his video feed, and one for
screensharing, both supporting FEC, and with the video feed
configured for simulcast. Alice accepts these video flows, but
does not add video flows of her own, so they are handled as
recvonly. Alice also specifies a maximum video decoder
resolution.The SDP for |offer-B1| looks like:|offer-B1-candidate-1| looks like:|offer-B1-candidate-2| looks like:|offer-B1-candidate-3| looks like:The SDP for |answer-B1| looks like:|answer-B1-candidate-1| looks like:|answer-B1-candidate-2| looks like:|answer-B1-candidate-3| looks like:The SDP for |offer-B2| is shown below. In addition to the
new m= sections for video, both of which are offering FEC, and
one of which is offering simulcast, note the increment of the
version number in the o= line, changes to the c= line,
indicating the local candidate that was selected, and the
inclusion of gathered candidates as a=candidate lines.The SDP for |answer-B2| is shown below. In addition to the
acceptance of the video m= sections, the use of a=recvonly to
indicate one-way video, and the use of a=imageattr to limit the
received resolution, note the use of setup:passive to maintain
the existing DTLS roles.This example demonstrates the early warmup technique
described in
. Here, Alice's
endpoint sends an offer to Bob's endpoint to start an
audio/video call. Bob immediately responds with an answer that
accepts the audio/video m= sections, but marks them as sendonly
(from his perspective), meaning that Alice will not yet send
media. This allows the JSEP implementation to start negotiating
ICE and DTLS immediately. Bob's endpoint then prompts him to
answer the call, and when he does, his endpoint sends a second
offer which enables the audio and video m= sections, and
thereby bidirectional media transmission. The advantage of such
a flow is that as soon as the first answer is received, the
implementation can proceed with ICE and DTLS negotiation and
establish the session transport. If the transport setup
completes before the second offer is sent, then media can be
transmitted immediately by the callee immediately upon
answering the call, minimizing perceived post-dial-delay. The
second offer/answer exchange can also change the preferred
codecs or other session parameters.This example also makes use of the "relay" ICE candidate
policy described in
to minimize the ICE
gathering and checking needed.The SDP for |offer-C1| looks like:|offer-C1-candidate-1| looks like:The SDP for |answer-C1| looks like:|answer-C1-candidate-1| looks like:The SDP for |offer-C2| looks like:The SDP for |answer-C2| looks like:The IETF has published separate documents
describing the security
architecture for WebRTC as a whole. The remainder of this section
describes security considerations for this document.While formally the JSEP interface is an API, it is better to
think of it is an Internet protocol, with the application
JavaScript being untrustworthy from the perspective of the JSEP
implementation. Thus, the threat model of
applies. In particular, JavaScript can
call the API in any order and with any inputs, including
malicious ones. This is particularly relevant when we consider
the SDP which is passed to setLocalDescription(). While correct
API usage requires that the application pass in SDP which was
derived from createOffer() or createAnswer(), there is no
guarantee that applications do so. The JSEP implementation MUST
be prepared for the JavaScript to pass in bogus data instead.Conversely, the application programmer needs to be aware that
the JavaScript does not have complete control of endpoint
behavior. One case that bears particular mention is that editing
ICE candidates out of the SDP or suppressing trickled candidates
does not have the expected behavior: implementations will still
perform checks from those candidates even if they are not sent to
the other side. Thus, for instance, it is not possible to prevent
the remote peer from learning your public IP address by removing
server reflexive candidates. Applications which wish to conceal
their public IP address should instead configure the ICE agent to
use only relay candidates.This document requires no actions from IANA.Harald Alvestrand, Taylor Brandstetter, Suhas Nandakumar, and
Peter Thatcher provided significant text for this draft. Bernard
Aboba, Adam Bergkvist, Dan Burnett, Ben Campbell, Alissa Cooper,
Richard Ejzak, Stefan Hakansson, Ted Hardie, Christer Holmberg
Andrew Hutton, Randell Jesup, Matthew Kaufman, Anant Narayanan,
Adam Roach, Robert Sparks, Neil Stratford, Martin Thomson, Sean
Turner, and Magnus Westerlund all provided valuable feedback on
this proposal.WebRTC 1.0: Real-time Communication Between
BrowsersEricssonCiscoMozillaMicrosoft CorporationGoogle3rd Generation Partnership Project; Technical
Specification Group Services and System Aspects; IP
Multimedia Subsystem (IMS); Multimedia Telephony; Media
handling and interaction (Release 12)3GPP TS 26.114 V12.8.0For the syntax validation performed in
, the following list of ABNF
definitions is used:AttributeReferenceptime Section 9maxptime Section 9rtpmap Section 9recvonly Section 9sendrecv Section 9sendonly Section 9inactive Section 9framerate Section 9fmtp Section 9quality Section 9rtcp Section 2.1setup Sections 3, 4, and 5connection Sections 3, 4, and 5fingerprint Section 5rtcp-fb Section 4.2candidate Section 15.1remote-candidates Section 15.2ice-lite Section 15.3ice-ufrag Section 15.4ice-pwd Section 15.4ice-options Section 15.5extmap Section 7mid Sections 4 and 5group Sections 4 and 5imageattr Section 3.1extmap (encrypt option) Section 4msid Section 2rid Section 10simulcast Section 6.1tls-id Section 4Note to RFC Editor: Please remove this section before
publication.Changes in draft-24:Clarify that rounding is permitted when trying to maintain
aspect ratio.Update tls-id handling to match what is specified in
dtls-sdp.Changes in draft-23:Clarify rollback handling, and treat it similarly to other
setLocal/setRemote usages.Adopt a first-fit policy for handling multiple remote
a=imageattr attributes.Clarify that a session description with zero m= sections
is legal.Changes in draft-22:Clarify currentDirection versus direction.Correct session-id text so that it aligns with RFC
3264.Clarify that generated ICE candidate objects must have all
four fields.Make rollback work from any state besides stable and
regardless of whether setLocalDescription or
setRemoteDescription is used.Allow modifying SDP before sending or after receiving
either offers or answers (previously this was forbidden for
answers).Provide rationale for several design choices.Changes in draft-21:Change dtls-id to tls-id to match MMUSIC draft.Replace regular expression for proto field with a list and
clarify that the answer must exactly match the offer.Remove text about how to error check on setLocal because
local descriptions cannot be changed.Rework silence suppression support to always require that
both sides agree to silence suppression or none is used.Remove instructions to parse "a=ssrc-group".Allow the addition of new codecs in answers and in
subsequent offers.Clarify imageattr processing. Replace use of [x=0,y=0]
with direction indicators.Document when early media can occur.Fix ICE default port handling when bundle-only is
used.Forbid duplicating IDENTICAL/TRANSPORT attributes when you
are bundling.Clarify the number of components to gather when bundle is
involved.Explicitly state that PTs and SSRCs are to be used for
demuxing.Update guidance on "a=setup" line. This should now match
the MMUSIC draft.Update guidance on certificate/digest matching to conform
to RFC8122.Update examples.Changes in draft-20:Remove Appendix-B.Changes in draft-19:Examples are now machine-generated for correctness, and
use IETF-approved example IP addresses.Add early transport warmup example, and add missing
attributes to existing examples.Only send "a=rtcp-mux-only" and "a=bundle-only" on new m=
sections.Update references.Add coverage of a=identity.Explain the lipsync group algorithm more thoroughly.Remove unnecessary list of MTI specs.Allow codecs which weren't offered to appear in answers
and which weren't selected to appear in subsequent
offers.Codec preferences now are applied on both initial and
subsequent offers and answers.Clarify a=msid handling for recvonly m= sections.Clarify behavior of attributes for bundle-only data
channels.Allow media attributes to appear in data m= sections when
all the media m= sections are bundle-only.Use consistent terminology for JSEP implementations.Describe how to handle failed API calls.Some cleanup on routing rules.Changes in draft-18:Update demux algorithm and move it to an appendix in
preparation for merging it into BUNDLE.Clarify why we can't handle an incoming offer to send
simulcast.Expand IceCandidate object text.Further document use of ICE candidate pool.Document removeTrack.Update requirements to only accept the last generated
offer/answer as an argument to setLocalDescription.Allow round pixels.Fix code around default timing when AVPF is not
specified.Clean up terminology around m= line and m=section.Provide a more realistic example for minimum decoder
capabilities.Document behavior when rtcp-mux policy is require but
rtcp-mux attribute not provided.Expanded discussion of RtpSender and RtpReceiver.Add RtpTransceiver.currentDirection and document
setDirection.Require imageattr x=0, y=0 to indicate that there are no
valid resolutions.Require a privacy-preserving MID/RID construction.Require support for RFC 3556 bandwidth modifiers.Update maxptime description.Note that endpoints may encounter extra codecs in answers
and subsequent offers from non-JSEP peers.Update references.Changes in draft-17:Split createOffer and createAnswer sections to clearly
indicate attributes which always appear and which only appear
when not bundled into another m= section.Add descriptions of RtpTransceiver methods.Describe how to process RTCP feedback attributes.Clarify transceiver directions and their interaction with
3264.Describe setCodecPreferences.Update RTP demux algorithm. Include RTCP.Update requirements for when a=rtcp is included, limiting
to cases where it is needed for backward compatibility.Clarify SAR handling.Updated addTrack matching algorithm.Remove a=ssrc requirements.Handle a=setup in reoffers.Discuss how RTX/FEC should be handled.Discuss how telephone-event should be handled.Discuss how CN/DTX should be handled.Add missing references to ABNF table.Changes in draft-16:Update addIceCandidate to indicate ICE generation and
allow per-m= section end-of-candidates.Update fingerprint handling to use
draft-ietf-mmusic-4572-update.Update text around SDP processing of RTP header extensions
and payload formats.Add sections on simulcast, addTransceiver, and
createDataChannel.Clarify text to ensure that the session ID is a positive
63 bit integer.Clarify SDP processing for direction indication.Describe SDP processing for rtcp-mux-only.Specify how SDP session version in o= line.Require that when doing an re-offer, the capabilities of
the new session are mostly required to be a subset of the
previously negotiated session.Clarified ICE restart interaction with bundle-only.Remove support for changing SDP before calling
setLocalDescription.Specify algorithm for demuxing RTP based on MID, PT, and
SSRC.Clarify rules for rejecting m= lines when bundle policy is
balanced or max-bundle.Changes in draft-15:Clarify text around codecs offered in subsequent
transactions to refer to what's been negotiated.Rewrite LS handling text to indicate edge cases and that
we're living with them.Require that answerer reject m= lines when there are no
codecs in common.Enforce max-bundle on offer processing.Fix TIAS formula to handle bits vs. kilobits.Describe addTrack algorithm.Clean up references.Changes in draft-14:Added discussion of RtpTransceivers + RtpSenders +
RtpReceivers, and how they interact with
createOffer/createAnswer.Removed obsolete OfferToReceiveX options.Explained how addIceCandidate can be used for
end-of-candidates.Changes in draft-13:Clarified which SDP lines can be ignored.Clarified how to handle various received attributes.Revised how attributes should be generated for bundled m=
lines.Remove unused references.Remove text advocating use of unilateral PTs.Trigger an ICE restart even if the ICE candidate policy is
being made more strict.Remove the 'public' ICE candidate policy.Move open issues into GitHub issues.Split local/remote description accessors into
current/pending.Clarify a=imageattr handling.Add more detail on VoiceActivityDetection handling.Reference draft-shieh-rtcweb-ip-handling.Make it clear when an ICE restart should occur.Resolve changes needed in references.Remove MSID semantics.ice-options are now at session level.Default RTCP mux policy is now 'require'.Changes in draft-12:Filled in sections on applying local and remote
descriptions.Discussed downscaling and upscaling to fulfill imageattr
requirements.Updated what SDP can be modified by the application.Updated to latest datachannel SDP.Allowed multiple fingerprint lines.Switched back to IPv4 for dummy candidates.Added additional clarity on ICE default candidates.Changes in draft-11:Clarified handling of RTP CNAMEs.Updated what SDP lines should be processed or ignored.Specified how a=imageattr should be used.Changes in draft-10:Described video size negotiation with imageattr.Clarified rejection of sections that do not have
mux-only.Add handling of LS groupsChanges in draft-09:Don't return null for {local,remote}Description after
close().Changed TCP/TLS to UDP/DTLS in RTP profile names.Separate out bundle and mux policy.Added specific references to FEC mechanisms.Added canTrickle mechanism.Added section on subsequent answers and, answer
options.Added text defining set{Local,Remote}Description
behavior.Changes in draft-08:
Added new example section and removed old examples in
appendix.Fixed <proto> field handling.Added text describing a=rtcp attribute.Reworked handling of OfferToReceiveAudio and
OfferToReceiveVideo per discussion at IETF 90.Reworked trickle ICE handling and its impact on m= and c=
lines per discussion at interim.Added max-bundle-and-rtcp-mux policy.Added description of maxptime handling.Updated ICE candidate pool default to 0.Resolved open issues around AppID/receiver-ID.Reworked and expanded how changes to the ICE configuration
are handled.Some reference updates.Editorial clarification.Changes in draft-07:
Expanded discussion of VAD and Opus DTX.Added a security considerations section.Rewrote the section on modifying SDP to require
implementations to clearly indicate whether any given
modification is allowed.Clarified impact of IceRestart on CreateOffer in local-offer
state.Guidance on whether attributes should be defined at the
media level or the session level.Renamed "default" bundle policy to "balanced".Removed default ICE candidate pool size and clarify how it
works.Defined a canonical order for assignment of MSTs to m=
lines.Removed discussion of rehydration.Added Eric Rescorla as a draft editor.Cleaned up references.Editorial cleanupChanges in draft-06:
Reworked handling of m= line recycling.Added handling of BUNDLE and bundle-only.Clarified handling of rollback.Added text describing the ICE Candidate Pool and its
behavior.Allowed OfferToReceiveX to create multiple recvonly m=
sections.Changes in draft-05:
Fixed several issues identified in the createOffer/Answer
sections during document review.Updated references.Changes in draft-04:
Filled in sections on createOffer and createAnswer.Added SDP examples.Fixed references.Changes in draft-03:
Added text describing relationship to W3C specificationChanges in draft-02:
Converted from nroffRemoved comparisons to old approaches abandoned by the
working groupRemoved stuff that has moved to W3C specificationAlign SDP handling with W3C draftClarified section on forking.Changes in draft-01:
Added diagrams for architecture and state machine.Added sections on forking and rehydration.Clarified meaning of "pranswer" and "answer".Reworked how ICE restarts and media directions are
controlled.Added list of parameters that can be changed in a
description.Updated suggested API and examples to match latest
thinking.Suggested API and examples have been moved to an
appendix.Changes in draft -00:
Migrated from draft-uberti-rtcweb-jsep-02.