Re: History of video compression paper

It would be a massive undertaking, but even
circulating and maintaining a incomplete but public
list would be very helpful to a judge in making
his/her Markman ruling. This way, the list would not
have to be recreated at great expense by each
defendant. Such a list could aid in the making of a
patent-free or royalty-free open source codec. 

Prior art could be added to a patent's file wrapper,
so that when the defendant requests a copy of the
prosecution history, the evidence is sitting there. It
is also possible for a patent to be re-examined by the
PTO, in light of new prior art for example.

Something like 4 out of 5 patents are really invalid,
but the burden of proof is up to the defendant. 

H.26L is attempting to define a royalty-free baseline
video coder, and may succeed if the patent holders
agree to break down and allow it. Otherwise, it's a
war between the standards, free codecs like Ogg, and
Microsoft (which could easy stand up against the MPEG
patent holders if it wanted to). In any event, it will
soon be blatantly obvious to every one that the basic
tools of video compression that contribute the vast
bulk of the coding efficiency are either patent free,
or at least royalty free by now.

Some of the patent holders within the MPEG-LA
portfolio could be easily thrown out, which would be
embarrassing. Others could be exposed for having no
relevancy to decoders, or for covering the same claims
as other patents in the same portfolio.

$0.02/hr is simply extortion, especially when it is
built upon such a weak portfolio of petty patents that
contribute little coding gain, while others in
history, or the committee itself, actually did most of
the inventing.

Most of the MPEG-2 patents are focused on interlaced
video coding, which is rarely used in DVD movies for
example.

The best thing to do is back up each component of your
codec by citing prior art. 

--- Erik Walthinsen <[email protected]> wrote:
> On Thu, 18 Jul 2002, mpeg minus wrote:
> 
> > It proves that variable shape motion compensation
> with
> > 8x8 and 4x4 blocks was invented way back in 1975,
> and
> > subpel filters in 1979! The whole MPEG paradigm is
> now
> > more than 17 years old.
> >
> > Lots of prior art listed, and I suspect much more
> will
> > be revealed soon!!
> 
> What would it take to provide a database of
> otherwise patented algorithms
> that are quite clearly invalidated by prior art, set
> up such that if
> someone tries to assert patent "rights" against any
> open-source software
> that uses such math, they can be quickly dispatched
> and sent back with
> their tail between their legs?
> 
> I know there's a potential mine field involved in
> challenging a patent,
> does anyone know exactly what it takes in order to
> safely assume that a
> patent is invalid?  Does it require a patent
> attorney to go through the
> materials and make absolutely certain all the
> various claims are bogus?
> 
> If such a thing is possible, it would open up a lot
> of algorithms for use
> in an open codec.  This is good.
> 
>       Erik Walthinsen <[email protected]> -
> System Administrator
>         __
>        /                  GStreamer - The only way
> to stream!
>       |    | M E G A        *****
> http://gstreamer.net/ *****
>       _  /_
> 
> 
> 
> --- >8 ----
> List archives:  http://www.xiph.org/archives/
> Ogg project homepage: http://www.xiph.org/ogg/
> To unsubscribe from this list, send a message to
> '[email protected]'
> containing only the word 'unsubscribe' in the body. 
> No subject is needed.
> Unsubscribe messages sent to the list will be
ignored/filtered.


__________________________________________________
Do You Yahoo!?
Yahoo! Autos - Get free new car price quotes
http://autos.yahoo.com

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to '[email protected]'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.