[License-discuss] proposal to revise and slightly reorganize the OSI licensing pages

[Chris Travers]

On Fri, Jun 8, 2012 at 7:46 PM, Rick Moen <rick at linuxmafia.com> wrote:

>> True, which is why I have sought out law review articles and case law.
>>  I would think that a case like MySQL v. Nusphere if it came up today
>> would still be a case of first impression, would it not?  I haven't
>> yet found a case directly on point (the closest maybe being Gates
>> Rubber).  Maybe you have found a closer one.
>
> Micro Star v. FormGen, CAI v. Intel, Gates Rubber.  (Don't be a painfully
> literal-minded computerist and insist you can learn the key concepts of
> derivative works only from software cases, please.)
>
I just think computer games are problematic here.  The reason is that
computer games are far more heavily protected than utilitarian
software.  This leads to issues in computer games that don't occur in
utilitarian software.  My reason for excluding them is that I think
they appear deceptively similar.  For example Micro Star v. FormGen
makes more sense when looked at as a multimedia case than as a
software case.  CAI v. Altia (is that the case you meant?) is mostly
like Gates Rubber in this regard.

Nowhere in these do I see any indication that mere inclusion of one
work in another creates derivation.  Indeed if it did, it would be
hard to imagine any case where a collected work or a compiled work was
not derivative of its members.
>
>>  Copyright in US law protects expressive elements to the extent that
>> they are separable from functional elements.  You can't use copyright
>> to protect pure function. [...]
>
> Quite.  You know, I _have_ studied the subject.

As have I which may be why the discussion is at least somewhat
interesting.  But I find it helps to start at the beginning.
>
>>  The main US case of this sort that doesn't involve video games
>> (excluding video games because they are protected both as software and
>> as audio-visual works so that is not a clear parallel) is Gates Rubber
>> v. Bando (http://digital-law-online.info/cases/28PQ2D1503.htm) where
>> the 10th Circuit applied an abstraction/filtration/comparison test to
>> determine whether one piece of software was derived from another,
>> concluded the district court got things wrong and sent it back to
>> district court.
>
> The 'abstraction-filtration-comparison' three part test developed in CAI
> v. Intel and further elaborated in Gates Rubber is relevant to
> _non-literal_ copying of copyright-eligible elements.  NuSphere's case
> is rather more brazen, and adjudication wouldn't require that conceptual
> test at all.  It would not even be relevant:

But literal copying isn't what makes a derivative work a derivative
work.  The AFC test gets you to the question of what's protected and
whether the overall work falls under copyright law generally.  It
doesn't tell you entirely what category of protected work it then
falls under.

I assume we agree that Ubuntu Linux is not a derivative work of bash,
or X.org, or any of the other pieces that they include.  It's properly
a compiled work including these, not derivative of these pieces.  If
we agree there, then we have to agree that the ISO of Ubuntu is also
not a derivative work of these pieces either.  Copyright permission is
required under US law to distribute the pieces, but only essentially
anthology rights, not the right to produce derivative works.  Assuming
we agree here so far, great.  We now know that it is possible for one
work to include another without being derivative of that work and the
question is limited to whether that is different just because we are
talking about a single binary.

In the cases like Micro Star, the issue is that by installing these
new levels, you get a new audio-visual work and it transforms the
original expressive multi-media into something new.  Note this would
be the case regardless of the technical details of implementation.  I
would go further however and suggest that the key factors here, as
they would be in any other multi-media case.  But this isn't a case
where you have functionality considerations.  It's merely a case of
protecting artistic expression.

If you are looking to literal copying, I would suggest that even
there, the functional/artistic boundary gets a lot more attention than
is clear.  For example in cases where literal copying is required to
achieve interoperability, every case I can find allowed the copying to
go forward.  See Sony v. Connectix, Lexmark v. Static Control, Oracle
v. Google (the judge noted that the API definition lines had to be
reproduced exactly, so that's copying of literal elements).  In
general I would suggest that where it comes to interoperability
generally, courts seem to be unwilling to enforce copyright.

So in this view, if Nusphere distributed the Gemini engine compiled
against MySQL's own header files, and with instructions for the end
user to link it, this might be permitted even in the absence of
copyright license.  It doesn't raise the artistic issues present in
Midway or Micro Star.  Distributing the library statically linked into
the executable?  I really think that's in uncharted territory, but if
they could be distributed separately and linked by the end user, it
seems to me to be more of a case of a compiled work than a derivative
one.

So I don't think the case that this is as clear cut as it appears on
the surface.  One might be able to argue that the binary distributed
is transformative of MySQL's work and therefore derivative.  But I
don't think it's as simple as just what's included.  That gets you to
"it's protected by copyright" but not to "it's a derivative, rather
than a compiled, work."

Best Wishes,
Chris Travers