Re: [rtcweb] VP8 vs H.264 - the core issue

[Silvia Pfeiffer <[e-mail]>]

On Thu, Oct 24, 2013 at 8:05 AM, Basil Mohamed Gohar
<[e-mail]> wrote:
> On 10/23/2013 04:55 PM, cowwoc wrote:
>> Harald,
>>
>>     I think it is premature to imply that VP8 is royalty free. I have
>> nothing against the codec (it's great) but it's my understanding that
>> Google can't guarantee that someone else won't exercise IPR rights
>> against VP8 in the future. The best we can say is that H264 requires
>> royalties today and VP8 *might* require royalties in the future. H264
>> has a slight advantage in this space in that we have well-understood
>> licensing terms.
>>
>>     I just wanted to put that out there so there are no confusions in
>> the future.
>>
>> Gili
>
> Actually, this is exactly the kind of FUD that has stifled the adoption
> of VP8 and, before it, Theora and Vorbis, as universally-available
> multimedia format.  It serves only to confuse the issue further, as I
> will explain below.
>
> For starters, there is no evidence whatsoever that there is a viable IPR
> concern with VP8, but there exist baseless allegations.  In fact, what
> little doubt that there might have been one was settled by the agreement
> signed between Google and MPEG-LA [1] a short while ago, which resulted
> in MPEG-LA withdrawing their attempt a forming a patent pool for VP8
> altogether.  An attempt, I might add, that had little public activity
> save for its initial announcement once VP8 was being concerned for
> international standards.  In fact, the extremely generous terms of the
> agreement lend credence to the fact that there was little that existing
> that would have been enforceable.
>
> Furthermore, the fact that there is an existing licensing structure for
> H.264 give exactly zero assurances of protections from IPR claims,
> because not all licensors of H.264 technology are a member of the
> MPEG-LA patent pool agreement, and there have been numerous patent cases
> related to H.264 and other technologies thought to be covered by RAND
> and FRAND terms.
>
> Finally, the current patent and IPR landscape, at least in the US, and
> widely in other portions of the world, eliminates the possibility of
> something *never* being under a patent threat, due to the presence of
> patent trolls that actively wait for adoption as well the sheer
> magnitude of patents and the very ease with which patent legislation can
> be brought up (including for those already "covered" by existing patent
> pools, e.g., H.264).
>
> So, in actuality, H.264 holds no advantage over VP8 in this regard, and
> the claim that VP8 is a liability to use is not evidenced by any actual
> unique tangible threat to date.
>
> [1] http://blog.webmproject.org/2013/03/vp8-and-mpeg-la.html


On top of all this, it seems to me that
http://www.ietf.org/rfc/rfc3979.txt requires all IPR holders on a
technology that is made part of an RFC to disclose their IPR and sign
a patent disclosure: http://tools.ietf.org/html/rfc3905 . I think this
process is trivial for VP8, but will require lengthy delays for
sorting out for H.264. In the interest of the Internet Community,
given that both codecs provide comparable quality at comparable
bitrates, we need to choose what is best for the Internet community.
RFC3979 even states this explicitly:

" In all matters of Intellectual Property Rights, the intent is to
   benefit the Internet community and the public at large, while
   respecting the legitimate rights of others."

It seems clear that given that there is no substantial technical
difference between the two, given that the IRP situation is so much
cleaner for VP8, and that the only known IPR holder for VP8 (ever
after challenges) is Google who are providing a perpetual royalty-free
license (http://www.webmproject.org/license/bitstream/), the
preference of the Internet community must clearly lie with VP8.

I would be surprised if the IESG - who has to consider IPR rights when
approving an RFC for publication - wouldn't have to overrule any
decision made by this WG to choose H.264 over VP8.

RFC3979 states:
" In general, IETF working groups prefer technologies with no known IPR
   claims or, for technologies with claims against them, an offer of
   royalty-free licensing."


Best Regards,
Silvia.