GPL Issue

Chris Travers chris at
Wed Sep 22 20:19:22 UTC 2010

On Tue, Sep 21, 2010 at 1:27 PM,  <dtemeles at> wrote:

> I defy anyone to provide a US-based statutory reference or court case that
> stands for the proposition that there is such a beast as a "copyright
> license" that is not by definition a contract subject to state law.  The
> "bare license" theory as presented by Mr. Moglen and the FSF is a fiction.
>  Note how neither he nor anyone else at FSF has ever cited a copyright case
> or statute in support of the "bare license" argument.  They simply raise a
> hunting example (that, by the way, may not even be an accurate recitation of
> the concept in the terms of real property law).  The "bare license" peddlars
> wholly ignore the fact that the patent, copyright and trademark acts
> specifically designate that all patents, copyrights and trademarks have the
> characteristics of personal property, not real property.  The hunting
> license example is therefore inapt.

While there are such cases (Jacobsen v. Katz comes to mind), I
personally think they have the possibility of causing a great deal of
harm if a reasonable violation of contract can result in severe legal
penalties (think of the Lori Drew case and how the CFAA was applied to
violations of MySpace's terms of service).  I personally hope courts
will move towards contract remedies except where conduct is well
outside any penumbral shield of the license contract.  However, we
shall see.
> So, why the lack of citations to case or statutory law?  There is no such
> thing as a "bare license" in copyright law.  Under patent law, the courts
> have used the term "bare license" to reference a non-exclusive patent
> license - i.e., an agreement not to sue the licensee for use in accordance
> with the terms of the license agreement (or in other words, a contract that
> is subject to state law).  None of these courts, per my research, have ever
> suggested that there is a license that is not also a contract.

I don't know about that.  Does a mere user of GPL'd software provide
sufficient consideration to the licensor to make that a valid
contract?  Or do we just have a system of legally binding notices here
(equivalent to an "open for business" sign and a "wet floor" sign)?
(As a note, there is no way to use GPL'd software in violation of the
GPL, so therefore this matter cannot ever come before the court.)

Obviously if one distributes GPL'd software, there is consideration
because any changes may indeed be valuable and thus if the original
authors obtain those changes, the ability to use them royalty-free is

> Software licenses are by their very nature an agreement between the licensor
> and licensee.  In other words, they are a contract.

IANAL, but I thought a contract required consideration.  What
consideration passes from the licensee to the licensor if the licensee
is a mere user (not talking about distribution here, just use)?  Is it
weightier than a peppercorn?

> There is nothing magic
> about open source licenses vs. proprietary licenses.  Both grant rights to
> use code subject to the terms of the license.  Both are agreements.  Both
> are subject to enforcement under state contract law.

In cases where distribution occurs, I'd be inclined to suggest it is
contractual.  However, for mere use, I don't see it.  Maybe you can
point out what consideration I am missing.  Not that I think this
matters.  I think the notice that the program may be used for any
purpose, and that it comes without any implied warranty are binding
even in the absence of any contract.

IOW, I see a license like the GPL consisting of basically three parts:
1)  Notice that the software may freely be used for any purpose (not
2)  Notice that the software comes with no warranties expressed or
implied (not contractual)
3)  An offer for a distribution contract.  Distribution entails
accepting that contract, and living up to the contract does involve
proper consideration.

> Hopefully, a case will arise in
> the next several years that the Supreme Court can take up to resolve the
> issue.

> Why is the contract v. "bare license" issue relevant, you might ask?  I
> suggest that most open source licenses fail to address enforcement in a
> practical way.

This is a fair point.

> The open source community cannot continue to hang its hat on
> the unicorn of the "bare license" in the hope that it can use the hammer of
> infringement remedies if/when a license must be enforced against a party who
> breaches the terms of the license.

Here is where I think you are likely correct legally but miss the
point in terms of how open source communities operate and how cases
would likely come about.

In general, I think it's worth noting first that social controls are
often stronger than legal controls in FOSS.  For example, it may well
be that certain things are beyond the scope of the GPL v2 that Moglen
et all think are covered.  From a legal perspective if this reaches
the court, the court may well rule that the GPL is weaker than people
think.  However, to a large extent this doesn't matter as much as one
might think.  There are very real business costs in appearing to be a
bad community player and these are entirely independent of judicial

Secondly the sorts of GPL derivation cases which are most likely to
come before the court IMO are those where there is no colorable
argument that the behavior is permitted (wholesale distribution of
binary-only versions), rather than the trickier cases (like what
exactly the difference between a work "based on" the GPL software is
vs mere aggregation when it comes to linking and the like).

Also, a lot of the uncertainty about the GPL right now has the effect
of creating a large no man's land that people justifiably avoid
wandering into, esp. regarding linking etc.  Whether this is legally
correct or not, it seems to work.

> We should instead develop more realistic
> contract-based remedies that address core enforceability issues in a manner
> that matches the reality of open source development. Let's start with a
> simple issue (tongue in cheek) that should be addressed - who actually has
> the right to enforce a license in the typical open source project that has
> potentially dozens or hundreds of contributors?  Who is entitled to damages?
>  (the same questions arise in the infringement context as well, but we'll
> save that for another day).

Those are fair questions.  In a very large number of projects it is a
non-issue though.  There is either a corporation behind the project
(such as Apache, or PostgreSQL) or a single individual has done a
lion's share of the work.  I don't know about you but I would expect a
court in the US to apply standing requirements such that the
individual must be a major contributor to the project, and that some
guy who contributed three patches some time age wouldn't have
particularized injury sufficient to grant standing.  Where a
corporation governs a project, or where a single contributor does most
of the work, I would expect a court to only grant those parties
standing.  IANAL though so I could be missing something.,

However back to the original question:

I don't think the bare license (to use, not to distribute) is
necessarily a problem since the notice disclaiming warranty is just
that.  It's a notice.  It's like a "wet floor" sign.   It has an
impact whether or not the individual agrees to it or not.  It's merely
a fair warning that if one uses the software, one assumes all risks of
doing so.

More information about the License-discuss mailing list