Notes, March 23

Den 23. mars 2017 13:47, skrev Harald Alvestrand:
> Today's main focus (apart from mergefest): Mark issues as appropriate
> (or not) for April 4 interim.
> 
> We already have WebRTC-PC PR #1011 and #1032, and issues #1077, #849,
> #763, #526
> as well as mediacapture-main issue #434.
> 
> Mediacapture-main
> =================
> Pull requests
> -------------
> 
> Issues
> ------
> #380 Define restrictions on device-info permission. (alvestrand)
> #384 Should we use [EnforceRange] on min/max in constraints? (aboba)
> #385 ConstrainLong only concern positive values (aboba)
> Something's happening in WebIDL-land. Waiting.
> 
> #396 Explanation of "advanced" constraint sets is confusing (burnburn)
In progress.

> #400 Add text to extensibility for consumer of MST (burnburn)
Discussion on what it's about and where it belongs - but this spec is
the only thing that people who write new MST consusmers are pretty
certain to read.

> #428 Implementation Suggestions section needs review (stefhak)
> hta asked guidou to review.
> 
> #434 Disable user media by default in cross-origin iframes (stefhak)
Stefan will write a PR.

> #438 What does it mean that the argument to applyConstraints() i ()
> Need assignee - I think we know what the spec needs to say now.
Adam will write a PR.

> 
> WebRTC-PC
> =========
> Pull requests
> -------------
> #1011 Eliminate NetworkError ()
Interim. Left open.

> #1026 strawman text to show how unverified media would work ()
> No response to 16 feb comments. Waiting for MMUSIC discussion. April tag?

Let's not tag it until we have the renewed writeup.

> 
> #1032 maxSimulcastStreams attribute in RTCRtpCodecCapability ()
> Interim fodder.
> 
> #1082 Handling RTX in RTCRtpCodecCapabilities ()
> Seems ready.

sdpFmtpLine should be "not set" or "not present" - Taylor to update and
merge.

> 
> #1087 Update call flow in Section 11.6 ()
> Needs to be reviewed in final form :-) (aboba: unwise to use your master
> branch?)

Merged.
> 
> #1088 Make "candidate" non-nullable in addIceCandidate parameter  ()
> Merge.

Merged.

New PRs:

#1094: Check for "closed" - discussion on whether it's OK to leave
dangling promises.
Deemed acceptable. Merged.

#1096: Remove "set of receivers".
Adam to update according to Taylor's suggestion and merge.


> 
> Issues
> ------
> #526 NetworkError event is not defined and might not be needed (aboba)
> #661 Add informative table of all things that can cause negotiat (burnburn)
> #763 Handling of simulcast errors (aboba)
> #788 Clean up remaining uses of 'set of receivers' (adam-be)
> #842 Define JSEP terms like "bundle" and "bundle-policy" in term (burnburn)
> #849 Specify an AllowUnverifiedMedia RTCConfiguration property  (fluffy)
> #881 Align getAlgorithm return value with Web Crypto (alvestrand)
> #942 Meta: auto-publish changes to the spec (alvestrand)
> #964 Handing SDP with more than one identity  (martinthomson)
> #1020 Don't fire events on a closed peer connection (adam-be)
> #1021 get/setParameters does not have a parameter for packetizati ()
Waiting for PR. Might have to drop.

> #1044 Section 12.2.1.1: enum errorDetail definition (aboba)
PR needed. It's enum, so lower-case-dash.

> #1053 Section 11.6: Issue 6 ()
Closed by PR.

> #1073 Need to specify which members of the encodings in "sendEnco (taylor-b)
Remove [[sendEncodings]] - it seems like a halfbaked addition. Consult
pthatcher first.

> #1074 Mark Identity as feature at risk? ()
Dan to write PR.

> #1077 Candidate from onicecandidate event and addIceCandidate are ()
gone

> #1079 Handling RTX in RTCRtpCodecCapabilities (aboba)
gone

> #1084 Note At-Risk features at front of document ()
Dan will do this just before CR stage.

> #1085 RTCRtpContributingSource.audioLevel not guaranteed to be in ()
Tricky. Interim, Taylor writing up.

> #1086 Make legacy API optional to implement ()
Interim.

> #1089 Update for structured cloning changes in HTML ()
Seems reasonably simple. Adam to take.

> #1090 When should RTCRtpContributingSource#audioLevel be null? ()
Taylor will write a PR (and discuss at the interim). Seems
uncontroversial, but since we have a CSRC block already, it can go there.

> #1091 When exactly is an SSRC RTCRtpContributingSource object upd ()
Interim
> 

#1092 DTLS failures: Interim.

#1093 rtcpTransport: Yes, cover it with feature-at-risk for "negotiate"

#1095 links point to legacy - Adam is fixing.

Received on Thursday, 23 March 2017 15:58:29 UTC