Re: What exactly is Derivative ?
On Thu, 25 Mar 1999, Richard Braakman wrote:
> Paul Nathan Puri wrote:
> > In copyright law, there are no defined distinctions. For the most part
> > you must judge for yourself. The discussion about the GPL, LGPL, etc., is
> > outside the scope of copyright law, and is governed by Contract law.
> > Therefore, 'What is a derivative' is not a pertinent question if your
> > query concerns an author's rights under the license in question.
> Free software licenses, however, derive their force from copyright law.
> To quote from the GPL:
> 5. You are not required to accept this License, since you have not
> signed it. However, nothing else grants you permission to modify or
> distribute the Program or its derivative works. These actions are
> prohibited by law if you do not accept this License. Therefore, by
> modifying or distributing the Program (or any work based on the
> Program), you indicate your acceptance of this License to do so, and
> all its terms and conditions for copying, distributing or modifying
> the Program or works based on it.
> If my work is not a derivative of the GPL'd work, under copyright law,
> then that work's license has no bearing on mine.
However, the author of the GPLed work or the author of the GPL have the
right to change the meaning of 'derivative' to suit their own purposes.
A properly worded license would read: "'derivative' is defined by the
Copyright Act Section XXX" (don't have my code handy); or "'derivative'
for purposes of the GPL means any work of authorship that is linked to,
symbolically linked to, compiled against, this source code." (or something
to that effect).
The force of point is that GPL is only governed by Copyright law if the
owner of the license explicitly says so. For example, in the GPL of the
Linux kernel, Linus had specify that system calls would not be considered
linking against the kernel, and thus not 'derivative.' The word
derivative in Copyright law is very weak and quite overly broad. It is
designed this way so that courts can decide copyright issues on a case by
case ad hoc basis. In this instance, Linus narrowed the definition of
'derivative' for purposes of Linux under the GPL.
I think it would be entirely appropriate for the GPL to have an extensive
definitions section. Otherwise, individual coders can make their own
definitions. Remember, that legal definitions can mean what you want them
to. For example, in a contract where you "sell" a "widget" to me, the
contract could read like this: "Sell" for purposes of this contract means
'give,' 'gift-over,' 'donate,' etc. etc. "Widget" for purposes of this
contract means "writing," "written plan," "thing," "tool," "your Stars
Wars collectibles" etc., etc. And this is why contract law has the much
more importance to the GPL than does copyright law.
I would be very careful in relying on copyright law to predict outcomes.
It is nearly impossible (or quite difficult and involved). Contracts and
licenses can and are given specific meanings and scopes of power.
Copyrights are amorphous property rights. `
The holder of the copy of a GPLed app has a contract with the author to do
or not to do certain things according the the license.
The holder of the copyright in the copy owns the right to revoke the
license, change the license,
I've looked over the GPL, and it specifies "the 'Program,' below refers to
any such program or work, and a 'work based on the Program' means either
the Program or any derivative work under copyright law." Now I ask you,
are Windows and Apple desktop looks and feel 'derived' from Xerox? Is
Spaceballs derived from Star Wars? (if you've seen 'The Freshman') Does
the Marlon Brando scene render this film a 'derivative' of The Godfather?
I take 10 lines of code from 20 different programs and somehow make them
all work as one program which one is it a derivative of? Is it a
derivative, a compilation, or both?
In other words, to base an entire discussion on whether something is a
derivative of another work is much like conjecturing how many angels can
dance on the head of a pin, etc. Think about it... what does it mean to
say 'a work is based on another when it is a derivative'? Even to say,
'it contains part... and modifications...' is like arguing whether we all
come from Eve.
Courts enforce the copyright holders license. Where the copyright holder
licenses under the GPL, and the condition is that a derivative must be
GPLed as well, the court can interpret according to a broad application of
the meaning of derivative or a narrow one. It will depend on the authors
intentions and case law. The case law is a vague mess; but if the authors
intentions are not made clear the court will enforce the lincense in favor
of the least responsible party under the circumstances. Where the
derivative author evidences intention to steal or subvert, the court will
find in favor or the copyright holder. But where the derivative author
appears the victim of vague terms, the court will not seek to punish.
If I were to argue against the GPL in court, I would say, "Your honor this
license is clearly inoperable because it merely invokes the Copyright
Laws, and it is obvious that the author already has that at his or her
disposal. This case for violation of a license should be dismissed
because this generic license does not implicate the specific facts
surrounding this particular work of authorship. If the case is on for
infringement, then he or she should have sued for copyright infringement
and not license violations."
I think that the the GPL will be effective in forcing derivative authors
to make their source code available. However, I do not think that it will
require them to license under the GPL as well. That would be like forcing
someone to make contract, and that is like indentured servitude. Courts
do not enforce such agreements. "If you come on my land you hereby agree
to mow the lawn" etc.
Right now the GPL operates as a kind of industry 'gentlemen's agreement.'
It has a powerful social movement behind it, but I doubt the legal
efficacy of it. It is the social movement that has all the Blue Chips
clamoring for marketing opportunities. They are keeping their lawyers out
of it because they do not want them to piss in the swimming pool. It
could get nasty.
I think it is advisable for authors to add some lines to the GPL that
pertains to his or her specific work. Define the scope of derivative,
make some terms more explicit. Seek peer review of the license to be sure
it complies with the spirit of the GPL.
As an upshot, as time goes on and (god willing) legal challenges to the
GPL only crop up in a few years a least, courts are likely to look to
industry practice and community standards to define the terms of the GPL.
These practices and standards are pretty well documented and voiced even
today. My argument for the GPL would be: "Your honor the GPL represents
a social movement at the heart of the internet revolution. The people
that brought you the internet more or less brought you the GPL. Now it
has a strong backing by many industry professionals and is strictly
adhered to within its circles. The GPL is not only a license given to a
specific work, but it represents the standards of practice by those that
write and use GPLed code. To allow someone to circumvent and take
advantage of the GPL by usurping and closing the source code and hence the
rights of others to utilize the application is to protect the thief and
ignore a widely accepted and utilized industry practice. Lawyers set
their own Code of Professional Responsibility. The GPL is taking shape as
an analogous Code of Professional Responsibilty for software programmers.
This License has enormous social benefits and should be entirely protected
and enforced according is own letter and to industry standards set by the
Open Source Community, which now includes IBM, Apple, Novell, Sun,
Netscape, etc." That seems pretty good.
Now you see how this license can be skewed to appear good and bad. It
looks worse under copyright law. It looks better under contract law where
the meanings of the terms are defined by industry standards.
I will do some research (I'm writing my IP paper on this subject) on how
courts enforce software licenses of work upon which a derivative is made.
I will post my results...
> Richard Braakman
> To UNSUBSCRIBE, email to email@example.com
> with a subject of "unsubscribe". Trouble? Contact firstname.lastname@example.org