Re: Agenda, May 21, 2015

Input (hoping to save time at the actual meeting) on some items.

On 20/05/15 23:19, Harald Alvestrand wrote:
> Action items:
>
> - HTA: Follow up WebAudio group's issue (Discussion on list)
> - Stefan: Follow up HTML 5.1 definition of direct assignment (new PR in
> queue)

Right, there is a PR in the queue, and in addition I've asked the html 
wg to make some clarifications to html5.1. Regardless of what happens 
there I think we should merge something along the lines of the PR.

>
> Overarching issues: Linkchecker is broken for IETF references. Dom can't
> fix it this week; just comment out the linkchecker in travis.yml.
>
> Media capture and streams
> -------------------------
> Pull requests - 4 open
>
> - #165 defer to HTML5.1 for direct assigment - suggest close given #171
> - #168 Distinguish between links to the active attribute and
> active/inactive - Adam to take annevk's advice. Status?
> - #170 new error objects. Dan, status?
> - #171 Referring to HTML 5.1 - stefan will present
>
> Issues - 7 open
>
> - #118 Practical algorithm - add as appendix? Discuss on list?
> - #127 Cancellation of a GUM future - awaiting others' solutions
> - #157 naming of enumerateDevices - no action
> - #159 Active attribute - see PR #168
> - #161 Remove "Direct assigment" - see PR #171
> - #162 MediaStreamError should not be an interface - see PR #170
> - #163 Please use [Exposed] - list discussion, but no consensus to
> change from previous decision. Adam, status?
>
> Last Call comments - 19 open, 13 have no proposed resolution listed
>
> Want to have proposed resolutions by Friday if possible. Missing:
> - LC-3026 Internationalization of "label"
> - LC-3021 Constraint specification - assigned to Dan - suggest better
> pointers
> - LC-3022 Latency constraint - assigned to pthatcher
> - LC-3017 MediaStreamError - assigned to Dan
> - LC-3027 Internationalization of Error's "message" - assigned to Harald
> - LC-3025 onaddtrack firing when you call AddTrack - assigned to Adam -
> suggest no change
> - LC-3023 more info on audio devices - assigned to Harald
> - LC-3010 Rename enumerateDevices - assigned to Harald
> - LC-3024 Remove registry - assigned to Dan - suggest "no consensus to
> change"
> - LC-3012 active attribute - assigned to Adam
> - LC-3011 Tamper-free language - assigned to Adam - suggest "we have fixed"
> - LC-3014 remove "detach from source" - assigned to Adam - suggest "we
> have fixed"
> - LC-3019 WebIDL errors - assigned to Dan
>
>
> WebRTC-PC
> ---------
> Pull requests
> - #29 mediadiscarded - stefan, any response from Martin?
Not yet, just pinged.

> - #195 replaceTrack - Stefan, status?

My summary right now reads:
===========================

There is consensus for:
* sender.replaceTrack() (as opposed to e.g. setTrack) as API model
* The identifier can remain unchanged, i.e. for the sequence
	senderX = pc.addTrack(X);
	senderX.replaceTrack(Y);
	the track on the remote side would all the time have the same Id
* If the track switched to cannot be sent without a renegotiation, 
replaceTrack should just fail (throw?), and the original track (X) would 
continue to be sent. Unclear though how fast the fail can be detected by 
the UA.

Regarding the Id: There are proponents for dropping the .id attribute on 
MediaStreamTrack’s and only have it on sender-receiver pairs. (Makes 
some sense to me, but would mean changes to gUM which is in last call. 
Could possibly also break apps that depend on the MST Id.). The majority 
(?) seem to think that the track Id should be carried over to the remote 
track, and its value stays at the Id of the track used with pc.addTrack.

It seems that we have some unclarity related to MID and MSID, but it 
seems more related to addTrack/removeTrack and m-line recycling (and not 
really to replaceTrack).
==========================
I'd like feedback before taking it to the list.


> - #214 Certificate management API - I think this is baked enough for
> inclusion.
> - #215 Expiring certificates - would be happier if there had been
> discussion on the list.
> - #220 Muted state - Adam, WDYT? Addressing Justin's comment?
> - #222 OPerations array -> promise-chain - seems to have mutated into
> I-don't-know-what. Suggestions?

I'm working on a summary.

> - #225 Pending and current SDP - Adam, status? (Jan-Ivar has opinons
> about naming....)
> - #228 QoS API (really priority API) - Dom has added references. Merge?
>
> Issues
>
> - How can we bash these? (again: Not today....)
>
>

Received on Thursday, 21 May 2015 07:00:14 UTC