Re: PR 258 (Add RTCRtpParameters.codec, RTCRtpCodecParameters, etc.)

I fixed the typo.  Thanks for catching that.

I was hesitant to add a typedef to the spec, since there are no other
typedefs in it at all (the style seems out of place, especially since it
would only be used in one place).

I'm also hesitant to make it an octet, since in principle there's no reason
that an RtpSender couldn't send with PT > 127.  It's just a current
limitation of RTP.  For example, if one head a header extension for adding
another byte to the PTs, then one wouldn't need to limit the PT to an
octet.




On Fri, Jul 24, 2015 at 6:14 AM, Bernard Aboba <Bernard.Aboba@microsoft.com>
wrote:

> Some comments on 258:
> https://github.com/w3c/webrtc-pc/pull/258/files
>
> + <dt>unsigned short payloadType</td>
> s/td/dt/
>
> Also, in ORTC it is:
>
> typedef octet payloadtype;
>
>
>
>
> On Jul 24, 2015, at 15:02, Stefan Håkansson LK <
> stefan.lk.hakansson@ericsson.com> wrote:
>
> On 24/07/15 14:44, Peter Thatcher wrote:
>
> Following the PR from https://github.com/w3c/webrtc-pc/pull/228, I have
>
> changed RtpEncodingParameters.priority to an enum with very-low, low,
>
> medium, and high.
>
>
> This is kind of what we decided a long ago about priorities.  But I
>
> forgot about it when I wrote the PR for RtpEncodingParameters.priority
>
> and made it a double.  This is an update to that.
>
>
>
> Do we still have consensus for using an enum for priority?  Look at the
>
> PR to see how it looks.
>
>
> Which PR is it? Priority seems to be part of #234 and #241.
>
> Anyway, I have concerns with the part
>
>           <dt>double priority</dt>
> +          <dd>
> +            <p>
> +              Indicates the relative priority of this encoding, across
> +              all RtpSenders of a given PeerConnection.  When there is
> +              limited bandwidth available to a PeerConnection, higher
> +              prioirty encodings will be sent with more bandwidth, and
> +              lower priority encodings will be sent with less
> +              bandwidth.
>
> in combination with the upcoming "min" and "max" bitrate attributes. How
> should he UA act if they conflict (e.g. a very high "min" and a low
> priority)?
>
> #228 only points at RTCWEB-TRANSPORT and I think that document only
> talks about DSCP marking.
>
> Stefan
>
>
>