Re: [rtcweb] Proposal for H.263 baseline codec

["Paul E. Jones" <[e-mail]>]

Stephan,

> The most commonly cited timeline for a widely in use technology to be
> "save" from a patent viewpoint, based on equitable defenses such as laches
> (in the US) is six years.  In some countries of significant size, this
> time is longer, and in others, equitable defenses do not exist.  (Very
> briefly, and perhaps incorrectly put, those equitable defenses allow a
> defendant to argue successfully that a patent cannot be enforced as the
> right holder knew that the patent claim was likely being infringed, and
> did not enforce the patent.).

Recall that Unisys forced people to pay royalties for using the GIF file
format long after it became widely popular.  The company asserted IPR claims
on GIF since it used a compression algorithm to which it acquired the
rights, and it started doing so right near the end of the 20-year period
during which a patent is considered valid.

In short, I really do not think one should ever assume it might be safe to
use any technology.

> In addition, as I pointed out in the meeting, the use of a video codec
> created by a body such as MPEG or ITU-T SG16 has the advantage of that the
> patents of all participating players are available at least under
> Reasonable and Non Discriminatory (RAND) terms.  This may sound like a Bad
> Thing if you operate under a business model that prevents you to pay
> anything for patent licenses, but it is surely a Good Thing if you are
> willing to dish out a moderate amount of money for a license.  RAND
> recently has gotten teeth, not so much in terms of the monetary
> compensation aspect, but in terms of difficulty (if not unavailability) to
> obtain injunctive relive, among others.  H.26x and the MPEG standards
> benefit from RAND commitments, VP8, AFAIK, does not.

I believe your point here is perhaps worth even more consideration.

I really know nothing about the IPR that exists or might be claimed on VP8.
That said, I know there has such an incredible amount of work by so many
companies to produce H.264 that I would be utterly surprised to find that
VP8 does not infringe on something.  All of the technology that went into
H.264 represents only a subset of all of the IPR that exists in the video
coding space.

It's the rest of the IPR, a bunch of IPR owned by companies who actually
have significant investment in video coding technologies, that I believe
people should worry about.  Everyone who worked on H.264 did so as part of
an open standards process, as you mention above.  They spent a lot of time
and energy in the process, coming to an agreement to license the technology
that went into H.264.  They did not agree to license technology that did not
go into H.264.  So, if one of those participating companies were to
subsequently sue over some IPR used in VP8, I would not dare call them a
troll.

Trolls always exist and may even lay claims to H.264, but I suspect it would
be much harder for a troll to lay claim to H.264.  That codec spent years in
development with input from tons of people.  If a troll were to come in and
lay claim to some part of H.264, I suspect there would be several companies
that would stand up and beat them down.  After all, a claim against H.264
would not only present a problem for the defendant in a lawsuit, but would
be an issue for every company with IPR on H.264.  Further, no matter what
part of H.264 such a troll might decide to pick on, there are a number of
world-class engineers who (as you indicated about H.263) could probably name
the person or persons who contributed the section of text, algorithm,
procedure, etc., all of which is covered with a known license.

VP8 does not have the benefit of such significant peer review and
collaboration, nor does it benefit from any IPR licensing agreements from
legitimate companies holding tons of video coding-related IPR.

Paul