Mailing-List: contact legal-discuss-help@apache.org; run by ezmlm
Precedence: bulk
Received-SPF: pass (athena.apache.org: domain of sa3ruby@gmail.com designates
 209.85.146.179 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=message-id:date:from:sender:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references:x-google-sender-auth;
        b=d3CFsrfdxZ8lFTewxbbsjgY/UlvTgYBmfE3lmRuY4h8+cA9nNRT9CNUojMmXZrPNLlJ9Tx/yJ1gWDVBl7NO0KcOCx4nY5ac4Z1jgGU9blvD09X68YBnbMeT7+WNN9lJMd9iPpMHLdnlS5HAu+4FpPcBxFaAv3hKCobVy+sXFEOc=
Message-ID: <3d4032300804100438t6e85b1f5ie5d137501fd0e916@mail.gmail.com>
Date: Thu, 10 Apr 2008 07:38:35 -0400
From: "Sam Ruby" <rubys@apache.org>
Sender: sa3ruby@gmail.com
To: "Geir Magnusson Jr." <geirm@apache.org>
Subject: Re: Open Source, Patents and "Patent Exhaustion"
Cc: "Legal Discuss" <legal-discuss@apache.org>
In-Reply-To: <9DB1F968-3595-4B3D-8328-9C2CB2E31542@apache.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
References: <9DB1F968-3595-4B3D-8328-9C2CB2E31542@apache.org>

On Wed, Apr 9, 2008 at 11:13 PM, Jeffrey Thompson <jthom@us.ibm.com> wrote:
> Geir,
>     I'll take a stab at some initial responses to your questions, but
>  understand that, unlike in a programmer's brain, a in lawyer's brain a bit
>  has more that 2 possible values.  Nothing is ever 0 or 1, it's usally
>  somewhere between .05 and .95.

This is actually a response to Geir's note, but I liked the context
that Jeff's "somewhere between .05 and .95" gives to the picture that
I'm adopting it.

On Wed, Apr 9, 2008 at 7:27 PM, Geir Magnusson Jr. <geirm@apache.org> wrote:
> I'd like to get a clear understanding of the intersection of patents and
> open source software.  I know that when i discuss this subject with various
> members of the Apache community, I get different answers, and some that make
> no sense based on my understanding of patent law (which I admit is limited),
> and feel that I'm confusing what amounts to an "ideal vision" versus
> "pragmatic understanding based on case law" (or something) - aka "reality".
>
>  I've long ago decided that I'm not very smart, and that I tend to learn and
> communicate better with examples.  So I'd like to see if I can resolve my
> confusion by getting answers to a set of questions about a plausible but
> hypothetical situation.  In order to keep it simple, I'd like to just
> consider the patent grant in the Apache License, and leave out for now the
> grant in the ICLA.

While it is fair to construct such a thought scenario, among other
things doing so highlights why we often require an ICLA and even a
CCLA in such cases.  And the reason relates to Jeff's legal algebra.

If we run into a case where the Apache License only gives us an answer
with a confidence of 0.5, having an ICLA and a CCLA each giving us
additional 0.5's may increase the confidence to a 0.875.

You mention patent exhaustion in the subject line.  There also is the
concept of implicit patent licenses.  And there may be other
additional reasons why a court may determine that a patent holder has
given up their right to enforce their patent.

More specifically, if somebody voluntarily contributes a software
patent to a project which is covered by a strong copyleft license, are
they giving up the right to enforce the patent in the context of a
commercial distribution?  That seems unlikely (as in closer to 0.05
than to 0.95).  If, however, they grant the right to a software patent
under a license that is conditional only on a patent termination
clause, are they giving up the right to enforce the patent to works
which are substantially derived from, but not bit-for-bit identical to
the original work?  Very possibly.  And by very possibly I mean that
this act alone may not be sufficient to change a 0.05 into a 0.95, but
may be sufficient to take a 0.93 to a 0.95, which is the top of the
scale that Jeff defines.  :-)

>  So, let me start by presenting a simple use case to give context.  This is
> supposed to be just the standard situation contemplated by Section 3 of the
> Apache License :
>
>  "Intel contributes under the Apache License a patented sorting
>   algorithm for use in Harmony's garbage collector.  Intel owns
>   the patent to the sorting algorithm."
>
>  I believe the following are true - corrections welcome :
>
>  True Statement 1 : "Users of Apache Harmony have a license from Intel for
> the sorting algorithm"
>
>  True Statement 2 : "Intel provides a license for the sorting algorithm to
> users of Apache Harmony, so that users can make, use, sell, offer to sell,
> import and transfer copies of Apache Harmony"  (this may be a longer version
> of Statement 1)
>
>  True Statement 3 : "Apache Harmony users have a license to the sorting
> algorithm for all future versions of Apache Harmony - as modifications,
> additions and improvements are made to Apache Harmony, all users of Apache
> Harmony still get a license for the sorting algorithm from Intel"
>
>  Now my questions. Assume I have a copy of Apache Harmony.  And yes, they
> get loopy towards the end, so be patient.
>
>  Question 1
>  ----------
>  If I create a new program that incorporates the code from Apache Harmony
> that implements Intel's sorting algorithm and I contribute that program to
> Apache Harmony, do I have a license from Intel for the sorting algorithm to
> make, use, transfer etc my new program (either combined or separate from
> Apache Harmony) under a license of my choice?

0.95

>  Question 2
>  ----------
>  If I create a new program that incorporates the code from Apache Harmony
> that implements Intel's sorting algorithm that I do *NOT* contribute to
> Apache Harmony, do I have a license from Intel for the sorting algorithm to
> make, use, transfer etc my new program under a license of my choice?

0.95

>  Question 3
>  ----------
>  If I incorporate the code from Apache Harmony that implements Intel's
> sorting algorithm into another open source VM, say JamVM (assuming license
> compatibility), do users of the JamVM have a license from Intel for the
> sorting algorithm to make, use, transfer JamVM?

0.8.  The less recognizable that the implementation of the concepts
are to the original contribution, the less certain the answer will be.

>  Question 4
>  ----------
>  If I create a new program that has no code from Apache Harmony that
> implements Intel's sorting algorithm and I do *NOT* contribute that code to
> Apache Harmony, can I depend on Intel's patent license for the sorting
> algorithm so I can make, use, transfer etc my new program under a license of
> my choice?  (IOW, can I indemnify my users wrt Intel's sorting algorithm?)

0.05

>  Question 5
>  ----------
>  If I purchase from BEA a virtual machine that implements Intel's sorting
> algorithm and BEA doesn't offer me indemnification, and Intel sues BEA for
> patent infringement over their sorting algorithm, can Intel also sue me for
> patent infringement for using BEA's virtual machine even though I once
> downloaded a copy of Apache Harmony?

Trick question.  In the US, anybody can sue anybody for any reason.

>  Question 6
>  ----------
>  As a followup to question 5, could I sell BEA my patent license for Intel's
> algorithm to use in their defense in Intel's patent suit?

Mu.  Unless you can come up with a scenario whereby Intel would prefer
whatever value add BEA might have to offer in this scenario, BEA would
look silly trying to sell something that Intel can download for free
from the ASF.

>  Question 7
>  ----------
>  For the situation in question 5, assume I haven't downloaded Apache
> Harmony.  Once I'm notified of Intel's suit against me, can I claim I ceased
> infringing the sorting algorithm patent by downloading the Apache Harmony
> tarball?

0.2.  I don't believe that the actual act of download is relevant, but
the fact that Intel voluntarily made the code available under a
license that is only contingent on a patent termination clause in the
event of a lawsuit may adversely impact Intel's ability to
successfully pursue a claim of infringement.

That's also why clarity is important.  It is why we insist on a
arguably redundant ICLA and/or CCLA in many circumstances.  And by
insist, I mean that there are cases where we can and should decline to
accept a contribution unless we are clear that not only may we use the
contribution under the terms of the Apache License, but that our
licensees may too.

>  Thanks in advance.  I'd really appreciate if the questions were answered
> directly with a "yes" or a "no" if possible (with a reason provided) and
> only if not possible, "maybe", with statements illustrating why.  If a
> question needs to be clarified or simplified to better facilitate an answer
> or communicate what is my probable intent, please do so and answer, or ask
> and I will clarify.

I tried to give additional clarification to answers that weren't
either 0.05 or 0.95.

>  Thanks
>
>  geir

- Sam Ruby

---------------------------------------------------------------------
DISCLAIMER: Discussions on this list are informational and educational
only.  Statements made on this list are not privileged, do not
constitute legal advice, and do not necessarily reflect the opinions
and policies of the ASF.  See <http://www.apache.org/licenses/> for
official ASF policies and documents.
---------------------------------------------------------------------
To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
For additional commands, e-mail: legal-discuss-help@apache.org