[M4IF Discuss] On2 and IPR

[William J. Fulco]

Craig,
>>At 6:37 PM -0800 3/22/02, William J. Fulco wrote:

>>You'd be truly amazed how much 1980s technology STILL works and is
>>useful for video-compression
>>[SNIP]
>>don't give "1982
>>video compression technology" the short shrift so quickly :-)

> At 7:31 AM -0800 3/23/02, Craig Birkmaier wrote:
> Excellent point Bill.
>
> This looks like an effort to entrench the
> interests of the18 companies that made a massive investment in the
> MPEG-2 process, and have since used that standard to proliferate an
> SDTV paradigm around the world that protects their investments in
> interlace and "601."

I don't know if the reason is quite so nefarious! MUCH of what is developed
in this business is done in (often willing) ignorance of prior-art. From my
reading - there was for instance - a lot of work done in compression and
imaging by NASA and it's contractors (clearly bandwidth from deep-space
probes needs to be minimized) - that could _ARGUABLY_ "submarine"
some/most/all of the patents in today's video-compression fray  - but
really, no one really cares. There is (was) a world full of grad-students
that think this stuff up, write their thesis, publish some work and then go
get jobs in the "real world" (back then that meant going to work for the
cold-war defense biz often) - leaving their cool little thingies forgotten
in the stacks of university libraries.
Having been on several standards committees myself, all the actual work gets
done by just a couple of very smart people while everyone else is there to
mostly "keep tabs" on what's going on and to protect their firm's interests.
However, there is SO MUCH intellectual property in the last 50 years of
video-compression research that it is impossible for these very few
smart-people to either keep up to date (or out-of-date :-) with what some
grad-student published while an intern at JPL.... or even what NBC/Sarnoff
Labs was doodling with 40 years ago to compress The NBC Nightly News to fit
on a couple of T1-Lines from NYC to LA...
Granted, if someone wanted to nuke the current patents with prior-art, it
would be big fight - and there's really nothing by it for MPEG-2 - however,
MPEG-4 with it's current licensing structure being like MPEG-2 while it's
application being far different - that's another matter. Lucky (or no) that
the container can deal with multiple solutions to this problem :-)
++Bill
wjf   NetworkXXIII.com