Notes, December 10

Den 10. des. 2015 15:14, skrev Harald Alvestrand:
> Overarching issues:
> 
> <none>
> 
> Mediacapture-Main Pull requests:
> 
> #277 Reformat with enforced line wrap at 80 ()
> Still unlandable. Assigned to Dom.
> 
> #284 Clarify that constraint in overconstrainederror is defined i (burnburn)
> Seems reasonable?

Dan has a suggested text (18 mins ago).
Dan will apply his fix and merge it (manual merge).

> #286 Fix facingMode in Capabilities example ()
> Merge-time?

Merged.
> 
> #289  Fix constraint description and add capabilty description fo ()
> Looks right to me. Question is really a new issue.
> (I raised a related issue as #293 after reading this)

Merged. Adam will raise an issue about constraints that are known but
not supported - does it make sense to report them? What about in
capabilities / settings?

> #290 Use single double instead of same max and min for the aspect ()
> LGTM

Merged.

> 
> #291 Remove MediaStream active/inactive events (fixes: #283) ()
> LGTM. Nice example!

Merged.

> Mediacapture-main Issues:
> 
> #244 Explain how extensions are to be done (burnburn)
> New text has been out for a week, not much feedback, what I've seen is
> positive. PR?

Dan is working on the PR.

> #246 track.applyConstraints() doesn't do anything. (burnburn)
> Jan-Ivar seems MIA. Dan?

Dan will work on it (after extensibility)

> #248 Conflicting information about OverconstrainedError.constrain (burnburn)
> OK after #284?

Can be closed when #284 is merged.

> #267 Permission model for cross-origin iframes (adam-be)
> Double-keying (top level + iframe origin) seems to be the preferred option.

Waiting for resolution with webappsec

> #268 Iframe sandboxing options for gUM (adam-be)
> Are we closer to a resolution?

No news, really.
> 
> #283  Active state spec inconsistencies (adam-be)
> Cleared by #291

Closed.

> #288 Dealing with LC-3019 (burnburn)
> "If you can't compile what we write, it's because the compiler can't do
> what we want". Dan?

No update.
> 
> #292 How is an implementation expected to scale video? ()
> Don't think this is a spec matter, but want to check that others also
> think so.

Closing for the TF, there might be an encoding issue.

> #293 Is there any sense in having defaults for SupportedCapabilities?
> My opinion is clear :-)

Adam will take a look at the WebIDL spec.

> 
> Webrtc-PC Pull requests
> 
> We'll have to wing it through these - I'm too behind :-)
> 
> #317 Make RTCDataChannels Transferable (stefhak)

Still waiting for EKR. Poked Jesup.

> #407 Bringing back the RTCIceCandidate interface, with 'init' dic (adam-be)

Merged. Adam will s/InvalidParameter/TypeError as a separate step.

> #416 RTCRtpSender.replaceTrack() fixes ()

Fine as far as it goes. More issues need raising.

> #419 fix setParameters invalid webidl ()

Merged. The semantics of calling setParameters with an empty parameters
dictionary seems unlikely to let it succeed, though.

> #421 Require sdp in RTCSessionDescription{,Init} ()

Merged.

> #422 Remove confusing paragraph on fourth party interception ()

Merged, per list discussion.

> #423 Add specific references to JSEP where possible ()

Merged. Now let's see how it goes.

> #424 Distinguish states from their attribute representation ()

Looks good, but has conflicts. Adam will merge.

> #425 Refer to definition of networking task source ()

Looks good, Adam will merge.

> #428 Don't create a default stream in 'dispatch a receiver' steps ()

Merged.

> #429 Adding expires attribute to generateCertificate. ()

Merged.

> #430 Add maxFramerate knob for simulcast ()

Merged. Lots of "up to the implementation" here.

> #431 Clarification for getRemoteCertificates() ()

Aboba will find a reference for "DER encoding" that doesn't suck...

> #432 Update RTCIceTransportPolicy ()

We'll take this change. It's a TypeError-breaker when you send "public"
to a browser that doesn't understand it, which is Not a Good Thing.
Would be better to change the API to take a DOMString "with values from
the enum", that we do in a few other places.

#433 Use unsigned long ssrc in stats

Merged. We'll deal with webrtc-stats separately.

Top of the hour. We might start assigning bugs by fiat instead of
waiting to get to them...


> Webrtc-PC Issues
> 
> Need assignees from #338
> 
> #8 Need to add public to RTCIceTransports (alvestrand)
> #133 Streams that become isolated generate errors on PC ()
> #140 DataChannel.onerror callback needs an error argument specifi (burnburn)
> #142 "Hold" unspecified (aboba)
> #150 when closing, all outstanding actions are cancelled and thei (adam-be)
> #179 Potentially private IP addresses revealed (alvestrand)
> #227 .mid attribute - random or Null when MID is not present in S (stefhak)
> #230 Add support for WebRTC Data Channel in Workers ()
> #244 set{Local|Remote}Description is underspecified (alvestrand)
> #246 RTCIceCandidate's sdpMid and sdpMLineIndex members (adam-be)
> #253 Assurance that requests to IdP proxy originate from the user
> (martinthomson)
> #257 ICE Candidate should have accessors for protocol-relevant it
> (alvestrand)
> #265 Add Privacy/security review questions (dontcallmedom)
> #286 The "consumer of the session description” conflicts with ter (adam-be)
> #295 Guidance for extending objects vs extending Stats needed (alvestrand)
> #296 Debugging ICE problems needs more info (aboba)
> #305 Describe what happens when media changes (fluffy)
> #308 RTCRtpSender.get/setParameters is underspecified (fluffy)
> #316 Don't use Interface.attribute to refer to properties of inst (adam-be)
> #319 Work through error definitions (burnburn)
> #328 RTCBundlePolicy Enum section should not be non-normative (aboba)
> #329 "none" as a RTCIceCredentialType (aboba)
> #330 Get rid of RTCAnswerOptions? (stefhak)
> #331 Creating an ICE agent is not defined (adam-be)
> #332 Timing of ICE gathering  (adam-be)
> #335 DOMError is removed from the platform; use DOMException inst (burnburn)
> #336 No definition of "applying an ICE candidate" (adam-be)
> #337 Interfacing between WebRTC spec and JSEP (dontcallmedom)
> #338 specify what parts of the SDP can be changed between the cre ()
> #339 Parsing stun/turn URLs ()
> #348 RTCRtpSender section doesn't include the definition for RTCR ()
> #358 Remove mid from sender/receiver ()
> #359 Set a priori codec preferences ()
> #363 m-line match up doesn't work with RtpSender.mid or RtpTransc ()
> #364 Browser calls close() on failed rollback ()
> #369 addTrack's streams parameter is unused. ()
> #370 Add drop option for RTCDegradationPreference  ()
> #373 peerIdentity and certificates are unused in setConfiguration ()
> #375 true as default values for dictionary is bad practice ()
> #377 is `audioLevel` in `RTCRtpContributingSource` positive or ne (aboba)
> #378 `getRemoteCertificates()` is ill-defined ()
> #382 Rejecting unallowed changes to configuration ()
> #383 Keep event handler attribute descriptions simle (don't speci ()
> #384 should "none" be a recognized value in RTCIceTransportPolicy (aboba)
> #386 Certificates expiry? ()
> #388 Section 10.1 (Simple Peer-to-peer Example) should explicitly ()
> #389 Should have a "closed" RTCPeerConnectionState ()
> #390 Ambiguous type of `ssrc` identifier; DOMString or unsigned l ()
> #391 Remove DTMF tones A-D (aboba)
> #392 Language for onnegotiationneeded is incomplete/inconsistent. ()
> #395 getRemoteCertificates() should return a sequence of base64 D ()
> #399 RFC5245 says nothing about multiple TURN servers, but WebRTC ()
> #400 STUN checks via TURN servers really should not have peer-ref ()
> #406 impact of bundle and rtcp-mux on the number of Transports ()
> #408 Clarify component for ICETransport (aboba)
> #412 Framerate knob for simulcast ()
> #413 Describe what happens to transceivers at rollback (pthatcherg)
> #415 Add activateSender() method (pthatcherg)
> #420 RTCSessionDescription should not allow nullable sdp ()
> #427 No reference to where TypeError is defined  ()
>

Received on Thursday, 10 December 2015 16:02:01 UTC