Re: [rtcweb] opportunity cost (was MTI video codec, charter, RFC 3929)

On Sun, Nov 17, 2013 at 8:46 PM, Harald Alvestrand <harald@alvestrand.no> wrote:
> On 11/17/2013 10:43 PM, Varun Singh wrote:
>
> +1, to focus on engineering and operational issues.
>
>
> Since firewall traversal was deemed to be a subject so specialized we
> couldn't debate it on the main list, perhaps we could have an MTI-only
> sublist?
>
> I'll personally commit to reading every message on it, but I suspect there
> are people on this mailing list who would like not to.
>

if other codecs are being discussed (H.261/3, Theora), it may affect the
traffic model discussion in RMCAT , i.e., how much variability in
sending rates can be expected for a target rate and how quickly can the
codec respond to a new rate request.

If the expected behavior is very different from the discussions on
traffic models (in Vancouver), having those discussions helps us
understand the congestion control requirements better.

>
>
> On Nov 14, 2013, at 1:06, Bernard Aboba <bernard_aboba@hotmail.com> wrote:
>
> Keith Drage said:
>
> "Agree
>
> I am at the point where I would prefer to spend the meeting cycles getting
> things we can agree on, rather than where we seem to be at the moment with
> an issue where there are two clear camps and no real sign of a compromise.
>
> Ultimately the market will decide (and some parts of it probably have
> already decided - which is probably the reason for no progress).
>
> Keith"
>
> [BA] Well said. With most of the RTCWEB WG drafts either having completed
> WGLC or being candidates for WGLC by the end of the year,  with some elbow
> grease it seems very possible to move the bulk of the documents to IETF last
> call within a few months at most.   Polishing the RTCWEB document set would
> yield multiple benefits.  Not only would it get us closer to the goal of
> standardizing the WebRTC protocol stack, but also might well turn up an
> issue or two we haven't thought enough about. Also, once we move the
> protocol stack further along, we'll have more cycles to spend on operational
> issues (like monitoring concerns discussed in XRBLOCK), which currently
> limit the ability to deploy WebRTC at very large scale.   Unfortunately,
> we've been spending so much time on the MTI video codec debate that less
> glamorous (but ultimately much more important) engineering work is being
> neglected.
>
> This is all by way of seconding your point that there is a real opportunity
> cost to the never-ending, energy sapping MTI codec discussion.  Personally,
> I'd much rather redirect the work of the Internet Engineering Task Force
> RTCWEB WG away from amateur lawyering toward engineering where we actually
> have expertise and could potentially make a difference.
>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb
>
>
>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb
>
>
>
> --
> Surveillance is pervasive. Go Dark.
>
>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb
>

-- 
http://www.netlab.tkk.fi/~varun/