[SciPy-Dev] Procedure for new code submission (pyHHT)

Robert Kern robert.kern@gmail....
Fri Jun 8 15:11:31 CDT 2012

On Fri, Jun 8, 2012 at 9:39 PM, Jaidev Deshpande
<deshpande.jaidev@gmail.com> wrote:
> Hi,
> I agree that HHT should remain isolated for the moment. When it
> matures, the community can decide where it should reside.
> As for the IPR status of the HHT, I want to share the reply Dr Norden
> Huang gave when I asked him about it.
> "The NASA patent covers only the US.  You are free to do whatever outside US
> territory as long as you do not make money in the US.  Go ahead with your
> project, and good luck."

Dr Huang does not own the patent anymore. Ocean Tomo Federal Services
LLC does, unless if they've sold it off. I do not believe he can give
you a license to the patent. His opinion about what can be done
without acquiring a license to the patent is in conflict with US law.

> Now, I would assume that this exempts research based on the HHT from
> the IPR regulations, even in the US. A lot of people have been doing
> it for almost a decade now, and many were Americans. And I don't think
> anyone pays or gets paid for using the HHT algorithms for academic
> purposes.

Sorry, I'm afraid not. I believe the current law of the land was laid
out in Madey v. Duke which interprets "commercial use" *extremely*
broadly. Universities charge tuition and take in research grant money,
and using patented inventions without paying royalties would make a
university's program more attractive, both from students paying
tuition and granting agencies, than otherwise. That was enough of a
commercial nexus for that judge. It also draws on previous cases that
limit the experimental use defense to cases where the use is "solely
for amusement, to satisfy idle curiosity, or for strictly
philosophical inquiry". University research programs are anything but
idle curiosity, I'm afraid. The Supreme Court declined to review the
case, leaving it the controlling precedent nationwide, and it has not
been overruled to my knowledge.

Here is the decision, edited down to the juiciest bits:


> And anyhow, there are many implementations of HHT on
> matlabcentral.com, there are freely available R packages too.

Sure. You can write open source software that includes patented
inventions. Users of that software have the responsibility to
independently acquire a license for the patent. The fact that academic
users have long ignored that responsibility just shows their ignorance
of the law (and I don't entirely blame them since it's complicated);
it is not a reflection of the law itself.

Robert Kern

More information about the SciPy-Dev mailing list