RE: A thought about patents

["Brijesh Kumar" <bkumar at ennovatenetworks.com>]

Henning Schulzrinne [schulzrinne at cs.columbia.edu]> writes
>
> In looking in multimedia-related patents, I'm also utterly
> amazed by the
> complete lack of citation of published technical articles or related
> work (RFCs, Internet drafts, etc.). The problem with many patents is
> that if submitted as a technical paper at even the lowliest
> conference,
> it would be laughed out of the program.

I have been on internal patent evaluation committees of my previous
employers (that doesn't mean I support idea of software patents.).
I would like to clarify a bit on the subject.

Granting of patents only means that a person grated a particular patent
was first to make "a claim" about the novelty of an idea or technique
as far as the patent office knows on the basis of "previous claims submitted
to it.". It is impossible for them to evaluate every  claim  on its
scientific
or intrinsic merits (even then they do a fairly decent job). The patent
office
basically notarizes the claim, date of a claim and the person who made a
claim.
It doesn't guarantee any usefulness or scientific merits of a claim (though
the person making a claim has to justify usefulness of a claim). It doesn't
mean that there are no better ideas already published or known to others in
the field, or the patent is of any great use to others. How do you expect
them
to know just about every thing under the sun?

> Maybe we need a peer review
> (under suitable secrecy arrangements) of patent applications by real
> experts, not the one patent examiner that seems to have missed the
> existence of engineering libraries in his or her undergraduate
> education.

Remember, a patent office works within a legal framework and is not a
ACM/IEEE technical review committee. Only a US law can stop software
patents.

All said, software patents should be declared crime against software
engineers (if not against, humanity.). The problem is that no big company
is likely to lobby for banning software patents because of their
vested interests since they already have a number of patents in their
portfolios.

Cheers,

--brijesh