Subject: RE: GPLv3 section 11
From: "Lawrence Rosen" <lrosen@rosenlaw.com>
Date: Wed, 11 Oct 2006 11:07:17 -0700

Stephen Turnbull wrote:
> There is a Big Bag of Claims.  You (the Licensor) reach in a grab a
> handful; this is granted to you as your Small Bag of Claims, aka
> patent.  Each Claim is discrete.  There is no notion of "combination
> of Claims"; each Claim can be tested by itself for whether a
> particular actual device infringes, and if you test two Claims, the
> result is simply the pair of results for each individually.

The scenario we had painted in our paper was that inventors *add* to the bag
of claims, not take some out. But that aside.... You are right that "each
Claim can be tested by itself for whether a particular actual device
infringes."

> You now write an Original Work, and the copyright entirely resides in
> you.  You convey the Work to me under a License which grants
> permission to manipulate it in various ways, which also contains a
> Licensee's Bag of Claim Tickets.  Each Claim Ticket contains the name
> of a Claim you covenant not to assert.  I now create a Derivative Work
> based on your work, and distribute it.

Look more closely at the License. Assume it says this:

   2. Grant of Patent License. Licensor grants You a worldwide, 
   royalty-free, non-exclusive, sublicensable license, under 
   patent claims owned or controlled by the Licensor that are 
   embodied in the Original Work as furnished by the Licensor,
   for the duration of the patents, to make, use, sell, offer 
   for sale, have made, and import the Original Work and 
   Derivative Works. [OSL/AFL 3.0]

If the License you're thinking about contains different words, longer or
more sentences, definitions of such things as "essential patent claims" and
other complexities, ask its author or licensor if that license means the
same thing as the one sentence above. For now, I'm not going to speak of an
abstract or draft license; I'm going to answer your questions as if the
above Grant of Patent License controlled the transaction.
 
> Something makes you mad at me, and you sue me.  First, you claim that
> I have violated the License with respect to the text of the Original
> Work (ie, under copyright law).  The Court considers our arguments,
> and rules that I have not.

Good for the Court and for you! Courts don't usually award damages because
the plaintiff is "mad" over "something." Congratulations on not being in
breach of contract (i.e., in breach of the copyright license). 

> Second, you claim that I have infringed your Small Bag of Claims,
> which you pour out on the table.  The Court picks out several and
> tells you to put them back in the bag; they are not embodied in my
> Derivative Work.  Now I take my Licensee's Bag and pour my Tickets
> out.  I start matching them up with your Claims.

OK, now there's serious stuff! I've apparently filed a lawsuit alleging
patent infringement by you, identifying one or more of my patent claims that
are infringed in some product you made, used or sold. 

Let's assume actual infringement. There are two remaining factual questions:

1. Is your product that infringes an Original Work or Derivative Work as
defined in the above License? 

2. Are the actual patent claims you infringe embodied in the Original Work
as I gave it to you?

If the answer to both questions is yes, then you are licensed for my patent
claims in your product. License is a defense to infringement. You win!

Other of my patent claims and other of your products aren't licensed. If you
infringe any of my patent claims in any of your other products, or any other
of my patent claims in your Derivative Works, you lose!
 
> Some of my Tickets may not match anything; that doesn't matter.  If I
> can tag each of your Claims with a Ticket, then I win.  If there are
> untagged Claims left on the table, then you win.

I prefer to undertake the analytical process above. 

> (If you win, then we leave patent law and go to general commercial law
> or something like that for calculating damages and determining
> appropriate remedies, based on the particular Claims infringed and not
> exempted by the covenant.)

Under patent law, damages can be expensive and you also risk an injunction
that would prevent you from making, using or selling your product
altogether.
 
> My question: is this a usefully accurate model, at least to start with?

As modified, and with my choice of License.... :-)

/Larry Rosen

> -----Original Message-----
> From: stephen@xemacs.org [mailto:stephen@xemacs.org]
> Sent: Tuesday, October 10, 2006 8:17 PM
> To: lrosen@rosenlaw.com
> Cc: 'Free Software for Business'
> Subject: GPLv3 section 11
> 
> Lawrence Rosen writes:
> 
>  > I'm not typically a shrinking violet on here, and I'm not afraid
>  > this time to speak up.
> 
> I hope you haven't taken me to mean you're holding back.  Rather, I
> think you have a rather different purpose here from mine or Tom's
> (which are quite distinct, IMO).  This means that we're posting at
> cross purposes, and this confuses me; I understand why Tom and I
> differ, but I don't really understand what you're getting at or why
> you don't pick out answerable parts of the discussion---I just don't
> have the necessary knowledge, I guess.
> 
>  > However, I suggest that each topic we discuss be in its own
>  > appropriate thread rather than lumped in with "A Primer on
>  > Infringing Patents in Software." I've changed the topic on this
>  > email to reflect the questions you're actually asking below.
> 
> Thank you; I really should have done that myself.
> 
> Nonetheless, I think my question can be framed entirely in the
> abstract.
> 
> There is a Big Bag of Claims.  You (the Licensor) reach in a grab a
> handful; this is granted to you as your Small Bag of Claims, aka
> patent.  Each Claim is discrete.  There is no notion of "combination
> of Claims"; each Claim can be tested by itself for whether a
> particular actual device infringes, and if you test two Claims, the
> result is simply the pair of results for each individually.
> 
> You now write an Original Work, and the copyright entirely resides in
> you.  You convey the Work to me under a License which grants
> permission to manipulate it in various ways, which also contains a
> Licensee's Bag of Claim Tickets.  Each Claim Ticket contains the name
> of a Claim you covenant not to assert.  I now create a Derivative Work
> based on your work, and distribute it.
> 
> Something makes you mad at me, and you sue me.  First, you claim that
> I have violated the License with respect to the text of the Original
> Work (ie, under copyright law).  The Court considers our arguments,
> and rules that I have not.
> 
> Second, you claim that I have infringed your Small Bag of Claims,
> which you pour out on the table.  The Court picks out several and
> tells you to put them back in the bag; they are not embodied in my
> Derivative Work.  Now I take my Licensee's Bag and pour my Tickets
> out.  I start matching them up with your Claims.
> 
> Some of my Tickets may not match anything; that doesn't matter.  If I
> can tag each of your Claims with a Ticket, then I win.  If there are
> untagged Claims left on the table, then you win.
> 
> (If you win, then we leave patent law and go to general commercial law
> or something like that for calculating damages and determining
> appropriate remedies, based on the particular Claims infringed and not
> exempted by the covenant.)
> 
> My question: is this a usefully accurate model, at least to start with?