ietf
[Top] [All Lists]

Re: [tsvwg] Last Call: <draft-ietf-tsvwg-rtcweb-qos-15.txt> (DSCP and other packet markings for WebRTC QoS) to Proposed Standard

2016-04-15 09:48:14

On Apr 3, 2016, at 3:37 PM, Black, David 
<david(_dot_)black(_at_)emc(_dot_)com> wrote:

I see a couple of Magnus's points that appear to need additional text
in the draft:

[1] Flow Type in application-facing browser API:

o Flow Type: The browser provides this input as it knows if
the flow is audio, interactive video with or without audio,
non-interactive video with or without audio, or data.

[... snip ...]

The main issue here is that to me it was not clear that "Interactive
Video with or without audio" allows for setting these DSCP values also
for the RTP stream containing audio also. This, I do see a need for
clarification on.

Propose an additional sentence:
OLD
  o  Flow Type: The browser provides this input as it knows if the flow
     is audio, interactive video with or without audio, non-interactive
     video with or without audio, or data.
NEW
  o  Flow Type: The browser provides this input as it knows if the flow
     is audio, interactive video with or without audio, non-interactive
     video with or without audio, or data.  For audio that is associated
     with a video flow, the flow type of the associated video MAY
     be used instead of the associated audio type.

I hesitate to say anything stronger than "MAY" here.

Looks good. 


[2] What does "interactive" mean in an implementation?:

We could add something along lines of ..... Currently in WebRTC, media sent 
over RTP is assumed to be interactive while media streamed over HTTP is assumed 
not to be. Future WebRTC extensions could allow streamed media over RTP.



The issue is that this document is called: DSCP and other packet
markings for WebRTC QoS. Then this document define something that is not
immediately mappable onto what is being defined in the other WebRTC
specifications. That is why I am raising that there need to be more
clear coupling. If that coupling is to mostly happen in another
document, can we at least then have a proposal on the table for that
change to ensure that the result is understandable.

Well, this TSVWG draft is definitely not the right place for a discussion of
when a video flow is interactive or non-interactive - I hope we can agree
on that.

Beyond that, as Cullen (Jennings) is both an author of this document and
one of the chairs of the rtcweb WG, I'd suggest that he and/or the rtcweb
WG propose an appropriate location for discussion of when a video flow
is interactive or non-interactive.  This TSVWG draft would then have an
additional sentence added, e.g.,

      See [TBD] for further discussion of how to determine
      whether a video flow is interactive vs. non-interactive.

I believe that the added reference here ([TBD] above) would be normative.

Cullen?

That discussion happened long ago for WebRTC and we decided we did not need a 
JavaScript controls point in the WebRTC API to indicate if RTP was interactive 
or not. If people start doing streaming video over RTP we can come back and 
revisit this and trivially add an API to indicate that in the W3C WebRTC API. 
Part of what drove this decision is the likes of Netflix / ITunes / Youtube are 
not asking the browser vendors for streaming media over RTSP or RTP. They think 
HTTP works much better for this. Thus the browser vendors see no need for non 
interactive video over RTP. I agree with Magnus that this might change some day 
in the future but right now, I think it's close enough that everyone can live 
with it. 

I'm not OK in treating it like some open issue that is still in discussion that 
somehow holds up this spec - it's not. 


Thanks, --David (as document  shepherd)