Re: Formal Objection to One vendor, One Veto

On Thu, Jul 16, 2009 at 9:19 AM, Chris Wilson<cwilso@microsoft.com> wrote:
> Silvia Pfeiffer [mailto:silviapfeiffer1@gmail.com] wrote:
>>As far as I understand MPEG-LA's patent pool: it is a means to get a
>>license for the "patent thickets" surrounding MPEG standards and thus
>>a distribution scheme for royalties to companies that have a patent
>>claim in the MPEG standards. The Patent Pool is, however, in no way
>>shape or form providing any form of protection to its licensees wrt
>>new patent claims of companies that are not taking part in the "patent
>>thicket" (http://www.mpegla.com/aboutus.cfm).
>
> You're quite right; in fact, given my limited understanding of patents (IANAL, TINLA), there is no way I know of to provide any such protection.  At all.  There can always be an unknown patent, and the holder of that patent can sue anyone they like if they have not made licensing arrangements.

Yes, I think that's right. There is always a risk of submarine
patents. A standards body can only try to reduce this risk as much as
possible by making sure that all players in the patent space have
previous knowledge of the standard and are therefore required to step
forward. Such a situation was in fact the reason why both Qualcomm's
case and Alcatel-Lucent's case eventually failed in court.

>>I do not think that setting up a patent pool for Theora makes any sense at all:
>
> I agree with you.  I didn't word that quite strongly enough in my response; but I think the reason why I mentioned that is that based on others' feedback, I expect there WOULD be a suit similar to the Alcatel-Lucent one, and you'll note who was sued in that one.  Microsoft is certainly not the only one who implements MP3 support in their products.  Video has a lot of patents.

Google, Apple, and even Mozilla can also be targets of such lawsuits -
and Google & Mozilla have already made the step to support Ogg Theora.
Presumably did their homework on the patent front to determine the
risk. Unfortunately, it is not common for companies to publish such
information, but instead everyone has to do their homework themselves
and make their own informed decision. I still hope such a patent
research could be done by W3C though.


>>first -  the only known patents that Theora makes use of are the ones
>>owned by On2 and On2 have already provided royalty-free licenses to
>>these; so, it would be a very lonesome pool
>
> ...until such time as other patents might be discovered, in which case one would expect the pool would want to acquire them.  But that would change the game substantially around Ogg.

... or the encoders could get adapted to work around such a patent (if
it really existed). Since it's all built on open source, a rollout of
such fixes would be relatively painless.


>>I do think, however, that two things make sense:
>>
>>first - the analysis of the patent space surrounding Theora and
>>publication of a document containing the outcomes of such an analysis,
>
> I'm not sure you could do that exhaustive a patent search, but okay.

I agree - it's probably impossible to do that completely exhaustively.
But the point here is not to be complete in everything that is
possible in the patent space, but the issue is to reduce the risk (and
the perceived risk).

It may actually be possible to make an exhaustive list of all the
technical approaches that are used in Theora and reference a
publication (scientific, expired patent, or On2 licensed patent) that
covers that approach. This may also help address the risk of patents.


>>secondly - the preparation of adoption of Theora as part of a
>>standard, and as part of that preparation there needs to be a call for
>>any patent holders to step forward within a given time frame. If such
>>a call was widely distributed and the distribution documented, a court
>>would be in a position to reject any later patent infringement claims
>>where the patent holder had obvious knowledge about the request to
>>step forward and have knowingly ignored that request.
>
> I don't know if that would have any legal standing.  (I'm not saying it wouldn't - I'm saying I'll well out of my depth and cannot respond.)

IIUC, the knowledge about the standards were a large part of what
helped in the two cases that I cited in previous emails on this thread
to be thrown out of court. So, I would think that such an approach
would help courts dismiss submarine patent claims. But IANAL.


> Perhaps the best answer would be to mandate that any conformant UA has to support either Ogg/Theora OR MP4/H.264.  At least that would mean only two encodings would be necessary to get interoperability.

For compatibility reasons, it would be preferable to have a single
baseline codec, but what you are suggesting may indeed become the
de-facto standard, just looking at the way things are going. Two
codecs is not the end of the world though, just extra effort for
publishers.

Regards,
Silvia.

Received on Thursday, 16 July 2009 05:00:31 UTC