Subject: RE: Patent prior art database
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Sat, 05 Jul 2008 12:05:01 +0900

Lawrence Rosen writes:

 > My point is that you're taking too narrow a view of "offense" and "defense."
 > Companies use their patents in lots of ways to defend their turf and their
 > profits.
 > 
 > Perhaps you mean, "I will use this patent only as a counter-claim if you sue
 > me for infringement," which is what most of our licenses now say. But that's
 > too narrow a definition.

No, I mean "a *patent is defensive* if it is primarily useful to
establish my right to practice the claim, but not very useful in
blocking others from conducting their businesses."

I agree with you, that *assuming you already have a patent* there are
many ways to use it, and you should not divide your portfolio into
"defensive" and "offensive" patents.  Rather, you should look at your
own technologies, your rivals' technologies, your markets' needs, and
formulate your strategies flexibly.  

However, here we started the thread talking about the *purely*
defensive strategy of filing applications to create a body of prior
art.  The questions then become, "should we pursue a strategy of
applying for patents on 'inventions' that are as bogus as the worst
patents that are granted?"  "Should we pursue technologies that we
don't need because we've discovered them and can (at relatively
nominal expense) file patent applications for them?"  "If further
effort/expense is required to go from application stage to get the
patent granted, should we expend it?"

In the "management of engineering" literature these kinds of practices
are more or less advocated.  Keep notebooks of neat things you
discover.  If it might be patentable, describe it to a trustworthy
colleague or boss, date, sign, and get her to countersign the notebook
page(s).  Make an appointment with the company's patent attorney.

I could go on (and on) describing these practices ....  This kind of
thing is what makes IBM the 800 pound gorilla of IP that it is.  But
it's not cheap, and we've seen several people from "smallish"
businesses (like Red Hat!) say "we can't afford to induce our
developers to do that, and they definitely don't think of it as fun".
OTOH, I gather Google (a mere 500 pound gorilla at this point) does do
those things.

 > If we are going to take advantage of the defensive opportunities of patents,
 > sometimes we have to license them for money that we can then use to promote
 > open source.

Sure, but I don't think anybody but RMS and his clones would disagree
with that.  At least, nobody in these thread has done so, so far.