Re: [MMUSIC] Progressing the rid draft (RTP Payload Format Constraints)

[Peter Thatcher <[e-mail]>]

On Fri, Feb 5, 2016 at 1:36 PM, Adam Roach <[e-mail]> wrote:

> On 2/5/16 15:01, Peter Thatcher wrote:
>
> Having just re-read draft-ietf-mmusic-rid-02 from to top to bottom, I
> think it's in a very good state.  I only have minor readability and
> editorial comments where I think it could be improvements.
>
>
> Thanks!
>
>  On small matters of readability:
>
> - We use the term "RID RTP streams", which is then defined in another spec
> as "source RTP Streams".  Can we just call them "source RTP streams" here
> too and save the reader some indirection?
>
>
> But it's not exactly "a Source RTP Stream". It's "a Source RTP Rtream,
> *or* the non-redundant RTP Stream of a Dependent Stream," which is a bit
> too unwieldy to use all over the document.
>

Oh yes, I forgot about that little wrinkle in the taxonomy.

​


> ​
>

> This distinction between "Source RTP Stream" and "RID RTP Stream" is small
> but important. If you want to suggest a different term for this construct,
> I'm quite keen to have a more intuitive term. But I spent probably way too
> much time scratching my head over this without coming up with anything
> better than "RID RTP Stream".
>
> You can argue that RFC 7576 should have defined a term that we could just
> use here, and I would heartily agree. But it didn't, and we've got to work
> with the terms as published.
>

You're right.  Let's just s
tick
​with​
 "RID RTP Stream"
​.
​


>
>
>
> - We use the terms "constraints", "parameters" and even "constraining
> parameters", which I believe all mean the same thing.  Can we just pick one
> and save the reader some disambiguation?
>
>
> I'll make a pass to clean this up. I'm going to go with "constraint" so as
> to avoid confusion with ftmp parameters.
>

​Sounds good.
​


>
>
>
> On the open issues:
>
> 11.1 (declarative SDP): I agree with keeping the current text.
>
> 11.2 (bitrate definition): I agree with keeping the current text.
>
> 11.3 (escaping values): I agree with option #3 (let a future extension
> define it, if needed).
>
> 11.4 (max-width and max-height): Either way is fine to me (included or
> not).
>
> 11.5 (max-fps definition): I agree with keeping the current text.
>
>
> Thanks for explicitly commenting on the open issues. :)
>
> And finally, some very minor editoral things:
>
> - The sentence "All "rid" parameters MUST be registered ... Section 12"
> doesn't have a period at the end.
>
> - The sentence "The SDP given below skips few lines" should be "The SDP
> given below skips a few lines".
>
>
> Good catches. Thanks.
>
> /a
>
>