Summary of decisions - TPAC WG meeting

*

This is the summary of decisions reached at the TPAC meeting of the
WEBRTC WG in Lyon, October 2018.


This summary is telegraphic in style; it is meant to be read in
conjunction with the agenda, slides and minutes stored here:


https://www.w3.org/2011/04/webrtc/wiki/October_22-23_2018


All decisions are subject to review on the mailing list; for decisions
where no further comments are heard on the list, we will assume that the
decisions listed here stand.


For the two decisions where it says explicitly “ask the list” or “to be
verified on the list”, I will send out an explicit thread-starting mail;
for all other decisions, please reply to this message or start a new thread.


MediaCapture


  *

    Combined origins are gone

  *

    Media functions are [SecureContext]


Screenshare

  *

    Decided

      o

        Fullscreen is info, not normative language

      o

        Spec for constraints will be merged

      o

        We will not allow URL white/blacklists in screenshare

      o

        Propose removing the “surface type” constraint

      o

        Remove MUST NOT on change sources

      o

        Move getDisplayMedia to mediaDevices

  *

    Not decided

      o

        Exclude own audio - maybe “echoCancellation”?

WebRTC-PC

  *

    Stop saying anything about track.id mapping

  *

    Rename transport.transport to transport.iceTransport

  *

    Set [[direction]] when SRD(answer)

  *

    Clarifications to codec parameters accepted

  *

    SetCodecPreferences remains on transceiver. Orphis to write tests.

  *

    Use order in answer as preference

  *

    Make codecPayloadType read-only (may be reconsidered)

Non-decisions:

  *

    SelectedCandidatePair needs further work (do candidates go away?)

  *

    codecPayloadType in addTransceiver: further discussion (hbos/aboba)

Testing

  *

    We will have a regular WPT review meeting

      o

        Chairs, editors, Youenn and Karthik have volunteered

  *

    We will create per-spec WPT directories

  *

    We will aim for a simulcast VI in January

  *

    We will aim for a hackathon in March (IETF)


NV use cases

  *

    Adopt document as WG document

WebRTC-ICE

  *

    We can change iceCandidatePair to interface

  *

    Peter will consult with CDN folks on forking

  *

    We will use NetworkInformation to represent network information (not
    define our own type)


WebRTC-SVC

  *

    We will adopt this document now.

Audio API

  *

    Consult with WebAudio WG

  *

    Consult with ML CG about what they need

Use of Streams

  *

    Decision: New APIs need to be Streams-based

  *

    Non-decision: either adopt stream-messages or need more info

  *

    Decision: Need advice from Streams folks about how to get processing
    off the main thread in the best way. <AI: Peter>

Datachannel

  *

    We will add an “onclosing” event.

Stats

  *

    STUN server stats acceptable.

  *

    Add .mid to senders and receivers.

  *

    Add streamIdentfiers to senders, obsolete stream stats object

  *

    Stats stay around except when ended (Jan-Ivar to write PR)

  *

    Do not remove “track” stats at this time. It’s not mandatory.

  *

    Expose IP addresses only if and as seen in addIceCandidate()

  *

    Copy size stats to outbound-rtp. Remove on sender.Discuss details on PR.


No consensus:

  *

    Leave networkType in the spec or take it out.

Workers


  *

    We wish to pursue Option 2 - Datachannel in service workers based on
    webrtc-ice.Youenn offers to specify this when there’s developer
    interest.

  *

    This work has lower priority than WebRTC 1.0 and webrtc-ice.

End-to-end Encryption (E2EE)

  *

    The WG adopts the E2E use case where we trust the application, but
    not the relay. (to be verified on the list)

  *

    The WG is still considering the generic “access to encoded data” use
    case.

  *

    The WG recommends basing API on TransformStreams

QUIC / Bring Your Own Transport

  *

    The WG will ask the list if we should adopt the WEBRTC-QUIC API
    document (in room: 2 opposed, ~10 in favor)

  *

    The WG encourages turning the “streams-based” slideware into a
    proposal. It doesn’t take a position on adoption yet.


Things that need doing

  *

    WPT process needs fixing

  *

    Need implementations to drive testing

  *

    Simulcast needs focus

      o

        Stats for simulcast - not specified

      o

        RID - is anyone implementing it?

      o

        Cosmo is able to do simulcast testing with SFUs

      o

        Should we reconsider spec separation?










*