🚀 go-pugleaf

In article <1990Apr9.204220.8231@algor2.algorists.com>
   jeffrey@algor2.ALGORISTS.COM (Jeffrey Kegler) writes:
>   ...
> It is not up to someone using the patented technology to
>challenge in court.  The most frequent strategy, and apparently best
>from a legal standpoint is to simply use the technology whose
>patentability you doubt, and put the burden of initiating legal action
>on the patent holder.  Over half of all patents granted by the Patent
>Office are tossed out by the court (and this figure includes a vast
>majority of patents on much more well-established grounds than the one
>above.)

I think what's meant here is that over half the plaintiffs in patent
infringement cases which actually go to trial are unsuccessful.  Most
patents are not tested in court, though, so it's not really accurate to
say that half the patents granted are tossed out.  Also, it frequently
happens that the court finds the patent to be valid, but uninfringed
(ie, the patent owner loses not because his patent is no good, but
because it doesn't cover the alleged infringing device or process).
A lot of potential cases don't make it to trial, either because the
patent is too weak (ie, the patent owner doesn't think it's worth the
expense of going through with a suit) or it's too strong (ie, the
alleged infringer choses to pay royalties rather than defend the suit).
In both such cases, the commercial value of the patent is also a
significant factor (ie, litigation is more likely in either case if
the royalties are high).
>
>Why bother with a Patent Office then?  It forces the patent holder to
>follow a standard form, creates a central reference for those worried
>about potential infringement to consult, and tosses out the most
>ridiculous claims.  (This used to include all software patents in
>happy days of yore).  By allowing software patents the Patent Office
>is not saying software is patentable.  It is saying that it does not
>want to risk being sued for denying the patent application, and has
>decided to allow the affected parties to contest the issue in the
>courts.

Courts can and do invalidate patents, but they do not do so lightly.
There is a strong presumption by the court that the patent office was
correct in deciding to grant a patent in the first place.  And it was
not the patent office, but rather the courts, which were the prime
force in bringing software into the realm of patentable subject matter.
That is, it was only in response to a decision of the US Supreme Court
(which only rarely hears patent cases) that the patent office revised
its hardline stance against software patents.
>
>Some software patents have made money from people who have decided
>paying royalties is easier than contesting the lawsuit, and many
>people have avoided using a patented technology for similar reasons,
>but I know of no software patent which has ever been tested in court.
>And note that most of those going to court with intellectual property
>claims in software these days are using the copyright laws, not the
>patent laws.  That tells you something.
>
Even the recent Supreme Court cases don't hold that software alone
constitutes patentable subject matter.  To be patentable, a program
must be claimed as a process for controlling a computer.  If, by
"software patent" Jeffrey means one with claims that refer to just
software in the abstract, it may be true that none have been tested
in court, but I doubt the patent office has ever granted such patents.
This is not the real issue, though, which dictates the choice between
patent and copyright protection for software.  More important is cost
(just the filing, issue and maintainance fees add up to over a thousand
dollars for a single patent, and attorney's fees are much more, while
the cost of registering a copyright is only $10) and time (besides the
time it takes to prepare a patent application, it typically takes a
year or more from the time the application is filed for a patent to
be issued.)  And there's no guarantee that a patent, once applied for,
will meet the standards of novelty, usefulness and nonobviousness
(even assuming it comprises patentable subject matter) required for
patentability.

John P. Kozma
kozma@comus.cs.tulane.edu