Re: Alternative decision process in RTCWeb
2013-11-30 09:57:10
Correct me if I am mistaken. I think the following two messages say is:
(1) just about everyone has IMPLEMENTED both H.264 and VP8
(2) people are worried about the availability or lack thereof of H.264
free licensing
(3) people are [should be] worried about lawsuits resulting from the
use of H.264 or VP8
Frankly, I am now fully in Sam’s camp:
On Nov 28, 2013, at 10:35 AM, Sam Hartman <hartmans-ietf(_at_)mit(_dot_)edu>
wrote:
I'd personally favor coin-flips, external arbitors or similar over
voting. We don't want to encourage voting because we could get into
situations where people block the consensus process in order to force a
vote. Absent things like well-defined membership and procedures to
avoid one company stuffing the ballot box, going there seems
problematic.
Coin flips are nice. They really create pressure among anyone who
actually cares about the outcome to compromise, to explore whether a
consensus can be built. However for the frequent situations where a
decision is critical but it really doesn't matter (even if some people
think it does), they get the job done.
We are way beyond picking a technically better solution. They smell like even
trade-offs. We are way beyond picking a businessly better solution. They are
both promised to be either free, nearly free, or
you-might-have-to-license-but-the-platform-you-run-on-already-has-a-paid-license-so-it-is-free.
We cannot tell which codec has less legal liability. Well, it is clear H.264
has more deterministic liability, but VP8’s liability today is literally
unbounded.
At this point, the King Solomon Solution
<http://en.wikipedia.org/wiki/Judgment_of_Solomon> is the solution. Cut the
baby in half. See who blinks. I need practice with my Tai QI sword - I will be
happy to provide the blade.
On Nov 28, 2013, at 6:05 PM, Cullen Jennings <fluffy(_at_)iii(_dot_)ca> wrote:
On Nov 28, 2013, at 3:45 PM, Phillip Hallam-Baker
<hallam(_at_)gmail(_dot_)com> wrote:
The issue here is different, the question is whether the people who write an
deploy code can come to a consensus. A vote is not going to make that
happen. If the WG votes for X but a browser provider with 30% of the market
insists that they won't do X, a vote is not going to change matters.
As a quick cheat sheet to where browser vendors might stand on this matter...
There are four major browsers. Firefox has said they will implement the
standard and everyone I have talked to in the WG believe they will regardless
of this video codec choice. Chrome awhile back said they will implement the
standards but they have not updated that opinion recently. Many people
believe they will implement the standard regardless of this choice (keep in
mind Chrome already has both 264 and VP8 in it for the video tag). IE and
Safari has not said if they will do WebRTC at all.
Mozilla plans to support both VP8 and H.264 and as far as I can tell prefers
Daala. Google prefers VP8. Apple and Microsoft prefer H.264.
From an IPR point of view. Mozilla plans to support both VP8 and H.264.
Chrome, IE, and Safari already use 264 and license the IPR. And Chrome also
uses VP8.
On Nov 29, 2013, at 1:22 PM, Stephan Wenger <stewe(_at_)stewe(_dot_)org> wrote:
Hi,
I agree with the basic sentiment of Philip’s email. However, in the (perhaps
forlorn) hope to shed some light into the IPR situation, allow me to correct
inline below some aspects of it that were painted in an overly broad brush.
I hope my characterization is reasonably impartial. All aspects below were
discussed on the rtcweb list ad nauseam, and in much greater detail.
Stephan
From: Phillip Hallam-Baker <hallam(_at_)gmail(_dot_)com>
Date: Thursday, 28 November, 2013 at 14:53
To: Dave Cridland <dave(_at_)cridland(_dot_)net>
Cc: IETF Discussion <ietf(_at_)ietf(_dot_)org>, Dave Crocker
<dcrocker(_at_)bbiw(_dot_)net>,
"rtcweb-chairs(_at_)tools(_dot_)ietf(_dot_)org"
<rtcweb-chairs(_at_)tools(_dot_)ietf(_dot_)org>
Subject: Re: Alternative decision process in RTCWeb
On Thu, Nov 28, 2013 at 10:04 AM, Dave Cridland
<dave(_at_)cridland(_dot_)net> wrote:
On Thu, Nov 28, 2013 at 2:40 PM, Dave Crocker <dhc(_at_)dcrocker(_dot_)net>
wrote:
BTW, as distasteful as it might be, is there a reason that making /both/ MTI
would not work?
[…]
The issues for vendors are litigation risk and cost.
If you are a commercial vendor with an existing H.264 license there is no
cost and no litigation risk for using that codec but the licenses you have
acquired are almost certainly specific to H.264. So using any other CODEC is
likely to create a substantial liability risk.
StW: A commercial vendor with an obtained pool license and current on your
reports and your bill has a license and insofar there is virtually no
litigation risk from the 30 pool members. I write “virtually no risk”,
because fraudulent lawsuits are not exactly unheard of, and there is nothing
one can do about those. The cost is $0.00 or $0.20 or $0.10 per codec,
depending on volume. According to Cisco, even without a license of your own,
the cost can be $0 if your product implements the download mechanism they
offer. Note that the definition of “codec” in the MPEG-LA agreements is
hardware centric and difficult to interpret in a software environment. So
there is cost (although potentially not for you).
Even with an MPEG-LA license, there also is litigation risk from those
rightholders who are not pool members, which is a not insignificant
percentage. However, the monetary damages that can reasonably be expected
from licensing or litigation are, according to recent rulings, some of which
are under appeal right now, quite low. This is because H.264 and its patents
were developed by a very broad community which agreed to license under RAND
terms. As there is not so much money to be made, the chance that a
rightholder would litigate outside of a strategic context (smartphone wars,
pool enforcement) is perhaps not overly high. Of course, that is of little
consequence once the multimillion $$$ bill from your lawyer rolls in. H.264
essential patents have been litigated at least 4 times since 2003, sometimes
as pool enforcement actions, and sometimes in the context of smartphone wars.
There are some 1200 licensees of the H.264 pool, including mega companies
and little guys one has never heard of.
In terms of patentable technologies, VP8 is considered by many (including
myself) mostly a subset of H.264. (It contains a few additional features
over H.264 but allow me to ignore those here.) Insofar, one can expect that
a subset of the H.264 essential patents may also ready on VP8, in addition to
a few VP8 specific patents. As far as licensing goes, Google has agreed to
compensate 11 rightholders (that also happen to be H.264 rightholders) an
undisclosed amount, and obtained a license from them which they generously
make available to us, along with a license under their own patents, and
subject to certain conditions which most people consider bearable. So I
think it’s safe to assume that these rightholders and google are not going to
litigate over VP8. As VP8 was not developed in an organization under RAND
IPR policy terms, there is no RAND safety net. Anyone other but the 12
companies above could theoretically sue over a VP8 patent and demand
sun/moon/stars and/or injunctive relief. Nokia has announced that they are
not willing to license patents they believe read on VP8, see here:
https://datatracker.ietf.org/ipr/2035/. Nokia has also asserted allegedly
VP8 essential patents against HTC in (at least) two distinct lawsuits in
Germany and one ITC complaint. One of the lawsuits is dismissed and under
appeal and the second is stayed pending an invalidation procedure (something
the German courts do only if they consider it likely that the patent is
invalid as granted). AFAIK, the ITC complaint is still in the procedural
warm-up phase. My personal take on these lawsuits is that they have to be
seen in the greater context of the smartphone patent wars. I may be wrong.
However, if I’m right, then there will probably be a settlement in the not
too distant future, with terms unpublished, the lawsuits go away, and we
would all be none the wiser. /StW
If you are an open source provider without a H.264 license the situation is
very different.
StW: Yes, it is different. H.264 implementations have been available in open
source form, since at least 2006. They have been part of major Linux
distributions forever. I’m following the litigation environment related to
video codecs quite closely, and I have not heard of a single lawsuit against
an open source developer or distributor regarding video codecs, including
H.264. All that is also true for VP8, except that VP8 is newer. /StW
This is not going to be settled by a vote. I am not speaking for any of the
parties but if I did have a dog in this fight I would have my corporate
counsel write a letter to the WG stating that we are not going to be bound on
the WG decision in this case.
StW: I don’t think that such a letter is needed. IETF RFCs, or parts
thereof, are regularly ignored by the implementer community, sometimes to the
detriment and sometimes to the benefit of the Internet at large. There’s no
standard’s police that could enforce compliant implementations. /StW
We are talking about a decision that could result in a hundred million dollar
lawsuit. The issue is not who is going to write code but who is likely to get
hit with a suit.
--
Website: http://hallambaker.com/
signature.asc
Description: Message signed with OpenPGP using GPGMail
|
|