Subject: RE: [CNI-(C)] Re: Open Source Licensing
From: "Lawrence E. Rosen" <lrosen@rosenlaw.com>
Date: Wed, 27 Aug 2003 11:31:32 -0700

[Cross-posted to CNI-COPYRIGHT and license-discuss@opensource.org]

Joseph Pietro Riolo wrote:
> I can see how you meant "perpetual" to be....
> 
> Still, that does not sit well with me.  How can lay people 
> and judges know that you meant that way without seeing your 
> post in CNI-COPYRIGHT?  Also, how many people who use your 
> license really understand your meaning of "perpetual"?

Who really cares what I meant?  I just drafted the contract; it is the
parties to the contract who must have a clear understanding -- a meeting
of the minds.

I hope that parties contracting under the Open Software License (OSL,
www.rosenlaw.com/osl2.0.html) really mean "perpetual" as "until the end
of time."  The fact that copyright law provides an escape mechanism to
protect against unequal bargaining power at the beginning of a copyright
term shouldn't prevent parties from saying "perpetual" when they
honestly, sincerely, presently intend it to be forever.

I am reminded of the "till death do us part" vows in marriage
ceremonies.  Are there any family law practitioners here who might
compare the legal effects of those words to the perpetual language in
software licenses?  

> Will it be better if you take out the word "perpetual" and 
> add another section on duration of license something like this:
> 
>      16. Duration.  This license shall be in force until
>      the expiration of copyright in original work or the
>      unequivocal and irrevocable dedication of the original
>      work to the public domain, whenever happens first.  In
>      no way this license will be terminated any time before
>      then.
> 
> That way, lay people and judges will understand that you
> do not mean the license to be overreaching, beyond the
> scope and term of copyright.

I was pleased to discover, after I published OSL version 2, that it was
a full page shorter than the GPL.  That fact alone increases the chances
that licensors and licensees will actually read it and try to understand
its implications and limitations.  While your proposed language would
indeed speak the truth, the cost of those words in terms of readability
and public understanding is prohibitive.  :-)  

Do the authors of proprietary software say such things in their
licenses?  Since the 35-year option-to-terminate rule (17 U.S.C. 203)
doesn't even apply to a work made for hire, is it practical to worry
about it in most software licenses?  Suppose we said, "this license
lasts until the copyright expires, and then you don't need a license at
all."  How would that help public understanding of the license?

But I'm still open to be convinced.  Others are encouraged to chime in
if they think a clearer Duration clause is needed in the Open Software
License.

> Earlier in the same post, you wrote:
> 
> > The law always trumps when it comes to issues of public 
> policy and it 
> > is unnecessary for a license to say otherwise.  ...
> 
> In light of the recent case between Harold L. Bowers and 
> Baystate Technologies, Inc., I have problem agreeing to your 
> assertion.

Bowers v. Baystate proves once again that "the law," as divined by the
courts, always trumps.  Just ask Baystate how painful it was to pay
damages!  In that case, the court reaffirmed that federal copyright law
does not necessarily preempt state contract law claims.  In particular,
the court held that a prohibition on reverse engineering in a
shrink-wrap license agreement is enforceable.  There's nothing new
there.  The court cites lots of precedent for that and the Supreme Court
denied certiorari.

Open source licenses, of course, unlike the proprietary license over
which the parties fought in Bowers, have a different objective, to
*guarantee freedom* rather than *impose restrictions* on the software.
A restriction on reverse engineering is impossible in an open source
license.

The OSL grants to licensees all the bundle of rights under copyright and
all the bundle of rights under patent, including the unlimited right to
use both the executable and source versions of the software and to
create derivative works from it.  There are only a few straightforward
conditions imposed on licensees, for example to protect the integrity of
the original author and to build a software commons through reciprocity
(what is sometimes pejoratively called "virality").  

Precisely because the OSL is an approved open source license, I want the
OSL to be strictly enforced under state contract law.  The Bowers case
helps ensure that my license will be interpreted according to its own
words rather than -- as with the GPL -- under the vague philosophy of
the Copyright Act.  To the extent that the OSL protects important policy
interests of the open source community and guarantees software freedom,
the license *should* prevail.  Only where the OSL might contradict a
public policy objective of the law, such as it would if I included a
provision that waived the right to terminate the license thirty-five
years later, should the law limit the right of parties to contract.  

/Larry Rosen
General counsel, Open Source Initiative (www.opensource.org)

--
license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3