Re: Proposed data channel API

Right, I was referring to the inclusion of the ACKing in the framework.

Agree that for SIP, it doesn't make sense. But I think there are other
cases where this functionality would be useful, and while it can always be
done by the application, having it in the framework can be much more
efficient since the ACKs can ride on the data flowing the opposite
direction. This could reduce message rate substantially.

Since it doesn't make sense for all applications, the use of this will be
optional. The action item for me is to flesh out exactly how this will work
and relevant use cases. If we can't find good use cases, it's probably not
worth the trouble.

On Wed, Nov 2, 2011 at 12:19 AM, Young, Milan <Milan.Young@nuance.com>wrote:

> Hello Justin,****
>
> ** **
>
> Just to be clear, the question wasn’t about the usefulness of ACKing
> unreliable sends.  The question was why that functionality needed to be
> included as part of the framework rather than detected through the generic
> messaging exposed by the framework.****
>
> ** **
>
> I was thinking about SIP in which, even on unreliable channels, not all
> messages have a traditional ACK.  So unless you were planning to explicitly
> generate an artificial ACK on ever message (an approach I find dubious),
> I’m not sure how this would work.****
>
> ** **
>
> Thanks****
>
> ** **
>
> ** **
>
> ** **
>
> ** **
>
> *From:* Justin Uberti [mailto:juberti@google.com]
> *Sent:* Tuesday, November 01, 2011 1:50 PM
> *To:* public-webrtc@w3.org
> *Cc:* Stefan Håkansson LK
> *Subject:* Re: Proposed data channel API****
>
> ** **
>
> The document at the link below has been updated with feedback from
> yesterday's discussion, specifically:****
>
> - initial section on security added****
>
> - minor changes to make DataStream more like WebSockets
> (sendMessage->send, added bufferedAmount)****
>
> - Added DataStream ctor****
>
> ** **
>
> Open issues:****
>
> - Use case for acking of unreliable sends****
>
> - onReadyToSend: WebSockets requires polling to determine when
> bufferedAmount == 0. The onReadyToSend callback provides a way to handle
> this more cleanly, and it's optional, so I'm reluctant to remove it.****
>
> On Mon, Oct 31, 2011 at 12:50 AM, Justin Uberti <juberti@google.com>
> wrote:****
>
>
> https://docs.google.com/document/pub?id=16csYCaHxIYP83DzCZJL7relQm2QNxT-qkay4-jLxoKA
> ****
>
> ** **
>
> A proposal for  <http://dev.w3.org/2011/webrtc/editor/webrtc.html>*
> http://dev.w3.org/2011/webrtc/editor/webrtc.html*, to discuss in
> tomorrow's TPAC meeting.****
> 5 The data stream****
>
> In addition to the MediaStreams defined earlier in this document, here we
> introduce the concept of DataStreams for PeerConnection. DataStreams are
> bidirectional p2p channels for real-time exchange of arbitrary application
> data in the form of datagrams. DataStreams can either be reliable, like
> TCP, or unreliable, like UDP, and have built-in congestion control, using a
> TCP-fair congestion control algorithm.****
>
> DataStreams are created via the new PeerConnection.createDataStream
> method. This method creates a new DataStream object, with specified "label"
> and "reliable" attributes; these attributes can not be changed after
> creation. DataStreams can then be added to a PeerConnection, much in the
> same way that MediaStreams are added. Since the semantics of the existing
> addStream API don't fit perfectly here (i.e. MediaStreamHints), we add the
> new addDataStream and removeDataStream APIs for this purpose. As with
> addStream/removeStream, these APIs update the internal session description
> of the PeerConnection, and cause a new offer to be generated and signaled
> through the PeerConnection signaling callback. Note that there is no
> requirement to add a MediaStream first before adding a DataStream; rather,
> it is expected that many uses of PeerConnection will be solely for
> application data exchange.****
>
> Like MediaStreams, multiple DataStreams can be multiplexed over a single
> PeerConnection. Each DataStream has a priority, which indicates what
> preference should be given to each DataStream when a flow-control state is
> entered. DataStreams with the highest priority are given the first
> notification and ability to send when flow control lifts.****
>
> In reliable mode, in-order delivery of messaging is guaranteed, which
> implies head-of-line blocking. In unreliable mode, messages may arrive out
> of order. In either mode, notifications of message delivery are
> communicated to the application via a callback; in unreliable mode,
> failures (defined as an elapsing of 2 RTT without an acknowledgement) are
> communicated through the same callback.****
>
> There is no maximum size to a datagram that can be sent over the data
> stream. However, messages are not interleaved on the wire, so a very large
> message will prevent other messages from being sent until its own send
> completes.****
>
> Encryption of the data stream is required. It is expected that
> applications that support DataStreams will support DTLS and DTLS-SRTP;
> while SDES-SRTP, or plain old RTP may be supported for legacy
> compatibility, there is no need to support DataStreams in these scenarios.
> ****
>
> In this draft, there is no inheritance relationship between MediaStream
> and DataStream, which is intentional due to the lack of a "is-a"
> relationship. However, it may make sense to hoist a Stream ancestor class
> for both MediaStream and DataStream, which would allow many of the existing
> PeerConnection APIs (localstreams, remotestreams, onaddstream,
> onremovestream, maybe removeStream) to refer to a Stream instead of the
> MediaStream that they currently reference. This would eliminate the need
> for "data" variants of all the aforementioned functions.****
> 5.1 Changes to PeerConnection****
>
> interface *PeerConnection* {****
>
>    [...]****
>
>    // Creates a data stream, either reliable or unreliable.****
>
>    // Reliability cannot be changed for a stream after it is created.****
>
>    DataStream createDataStream(in DOMString label, in boolean reliable);**
> **
>
>    // Adds a datastream to this PeerConnection. Will trigger new signaling.
> ****
>
> *   void add<http://dev.w3.org/2011/webrtc/editor/webrtc.html#widl-PeerConnection-addStream-void-MediaStream-stream-MediaStreamHints-hints>
> DataStream<http://dev.w3.org/2011/webrtc/editor/webrtc.html#widl-PeerConnection-addStream-void-MediaStream-stream-MediaStreamHints-hints> (in
> DataStream<http://dev.w3.org/2011/webrtc/editor/webrtc.html#idl-def-MediaStream>
>  stream);*****
>
>    // Removes a datastream from this PeerConnection. Will trigger new
> signaling.
>    void removeDataStream<http://dev.w3.org/2011/webrtc/editor/webrtc.html#widl-PeerConnection-removeStream-void-MediaStream-stream> (in
> DataStream<http://dev.w3.org/2011/webrtc/editor/webrtc.html#idl-def-MediaStream>
>  stream);****
>
>    [...]
> };****
> 5.2 The DataStream interface****
>
> interface DataStream {****
>
>    // Label, like MediaStream's |label|.****
>
>    // Maps to a lower-level stream identifier.
>    readonly attribute DOMString label;****
>
>    // Whether this stream has been configured as reliable.****
>
>    readonly attribute boolean reliable;****
>
>    // The relative priority of this stream.****
>
>    // If bandwidth is limited, higher priority streams get preference.****
>
>    // Default priority is zero.****
>
>    attribute long priority;****
>
>    // States, as in MediaStream.
>    const unsigned short LIVE = 1;****
>
>    const unsigned short ENDED = 2;
>    readonly attribute unsigned short readyState;
>    attribute Function onReadyStateChange;  ****
>
>    // Sends the supplied datagram.****
>
>    // Returns a nonnegative message id if the send was successful.****
>
>    // Returns -1 if the stream is flow-controlled.****
>
>    long sendMessage(in DOMString message);****
>
>    // Called when a message is received.****
>
>    // Arguments: DOMString message****
>
>    attribute Function onMessage;****
>
>    // Called when flow control lifts for this stream.****
>
>    // Arguments: None****
>
>    attribute Function onReadyToSend;****
>
>    // Called when a message has been delivered (or lost, if unreliable).**
> **
>
>    // Arguments: long id, bool success****
>
>    attribute Function onSendResult;
> }****
> 5.3 Example****
>
> // standard setup from existing example
> var local = new PeerConnection('TURNS  <http://example.net/>*example.net*',
> sendSignalingChannel);****
>
> // create and attach a data stream
> var aLocalDataStream = local.createDataStream("myChannel", false);****
>
> local.addDataStream(aLocalDataStream);****
>
> // outgoing SDP is dispatched, including a media block like:****
>
>     m=application 49200 <TBD> 127****
>
>     a=rtpmap:127 application/html-peer-connection-data****
>
> // this SDP is plugged into the remote onSignalingMessage, firing
> onAddStream****
>
> [remote] onAddStream(aRemoteDataStream);****
>
> // signaling completes, and the data stream goes active on both sides****
>
> [local] onReadyToSend();****
>
> [remote] onReadyToSend();****
>
>     // we start sending data on the data stream****
>
> var id = aLocalDataStream.send("foo");****
>
> // the message is delivered****
>
> [remote] onMessage("foo");****
>
> // the result is communicated back to the sender****
>
> [local] onSendResult(id, true);****
>
> // the data stream is discarded****
>
> local.removeDataStream(aLocalDataStream)****
>
> // new signaling is generated, resulting in onRemoveStream for the remote*
> ***
>
> [remote] onRemoveStream(aRemoteDataStream);****
> 5.4 Implementation notes****
>
> It is intended that this API map to the wire protocol and layering being
> defined in the IETF RTCWEB WG for the data channel. One current proposal
> for said protocol is****
>
> *http://www.ietf.org/id/draft-jesup-rtcweb-data-00.txt*, which is
> believed to match the requirements of this API.****
>
> ** **
>

Received on Wednesday, 2 November 2011 13:50:25 UTC