Re: RTCWeb proposal

[Harald Alvestrand <harald@xxxxxxxxxxxxx>]

On 12/05/2013 08:28 PM, Phillip Hallam-Baker wrote:

On Thu, Dec 5, 2013 at 7:12 AM, Harald Alvestrand <harald@xxxxxxxxxxxxx> wrote:

On 12/04/2013 06:11 PM, Phillip Hallam-Baker wrote:

On Wed, Dec 4, 2013 at 11:45 AM, Stephan Wenger <stewe@xxxxxxxxx> wrote:

Why MPEG-2?  MPEG-2 video is MPEG-1 video plus interlace support.  Do we want interlace?  More to the point, do we want to burden our decoders to implement interlace support (which is non-trivial) when no sane encoder would send interlace in a PC or mobile environment, even if the bitstream format would allow for it?

Stephan

As I said in a previous message, the actual choice of encoding does not really matter so much as the criteria.

The group is unable to come to agreement between the two choices. This is because both choices are de facto encumbered even if there may be dispute about the validity of the claims.

What I thought was a settled principle in the IETF is that the IETF does not choose a encumbered technology as MTI if there is a viable unencumbered alternative. 

H.264 is obviously encumbered so that is excluded from consideration. Anyone with deep pockets who attempts to use AV8 is likely to be sued and suffer substantial expense regardless of the merits of the case. Ergo AV8 is effectively encumbered.

Phil, if you can't even get the name of VP8 right, I think you are demonstrating a basic lack of knowledge about the area.

You are also proposing new solutions that seem to be minor variants of solutions already listed on the list of possible solutions ( http://tools.ietf.org/wg/rtcweb/trac/wiki) and doing so in a forum that is not the working group that has to come to consensus on any proposal.

Does it really matter what we call the codec if the issue that is blocking adoption is a well founded concern that it may be subject to a purported encumbrance?

I think the principle I was advancing was straightforward enough: Any decision to decide this matter by a vote is a matter for the whole IETF and not just the working group or the area ADs. And even if the IETF was to arrive at a consensus the decision would have no effect on interoperability.

Phil, if you had left it at that, i would not have reacted the way I did.

The specific reaction was to:

a) You suggesting that a codec was encumbered without bothering with enough fact-checking to get the name of the codec right.

b) You proposing specifically that MJPEG is a better solution to propose than the multiple maybe-royalty-free ones that have already been proposed in the WG (H.261, H.263 and Theora), without any justification for why you think this choice is either technically superior to or more IPR-proof than the others that have been proposed.

It was these two statements that caused me to use the term "uninformed punditry" - which Dave, probably correctly, thought was not part of the terminology we should use in IETF discussion.

I listen carefully when you hold forth in a pundit-like fashion on topics where I know you have worked for a long time, and where I know you know the terminology and technology better than I do.

This was not one of those times.