Re: Copyright lawyers analysis of Andreas Pour's Interpretation
On Tue, Feb 15, 2000 at 03:46:48AM -0500, Andreas Pour wrote:
> You fundamentally don't understand what a copyright in a collective work is.
> There are copyrights in the component works, and a separate copyright in
> their collection. Thus, any origniality used in kghostview (a component
> work) has no bearing on the originality used in the collective work (the
> selection of using kghostview together with Qt).
Creativity does not have to do with the number of distinct elements
in the collective work.
For example, I've seen collective works, such as books of short stories,
where only one or two of the short stories was under a different
copyright.
I think you're over-taken with your phone book analogy.
> Perhaps this is too abstract for you. Perhaps an example would help. I want
> to create an anthology of short fiction works. I select 10 short stories,
> each written by a different author, to include in it, based on an original
> selection criteria. There are now 11 copyrights in the anthology. Each
> author of the short stories retains a copyright in the 10 stories. I have a
> copyright in the selection of those stories. This means that if you want to
> come along and create an anthology that creates exactly the same 10 stories,
> even if you get rights to distribute the 10 individual works, you cannot
> combine them like I did (at least not by copying my selection -- if you come
> up with the combination independently, of course, copyright law does not
> restrict you). That combination is my copyright. My copyright does not
> extend to any part of the 10 stories -- read Section 201(c) again (the
> relevant language being, "Copyright in each separate contribution to a
> collective work is distinct from copyright in the collective work as a
> whole").
I already addressed this in the message to which you were replying,
where I said:
> > Once that work has been done, you're talking about copying and
> > distribution issues, not authorship.
Please don't confuse authorship decisions whith distribution decisions.
I don't think that there's any reasonable way that you can prove that
kghostsview was authored without the intent to create a collective work
which contains both libqt and ghostview:
If I were to play your analogy, the "short stories" here include:
ghostview, libqt, libkfile, libkfm, libkdecore, libXext, libX11, ...
And, significant creativity went into selecting which parts of these
works are used in the collective work, and significant creativity went
into the additional material which is unique to kghostview.
Finally, a court isn't going to care that, at some points, kghostview has
been distributed in pieces with instructions on how to build the work as
a whole, instead of solely as the work in its entirety. Why should it?
Traditional copyright cases typically involve far weaker evidence of
what the prior copyrighted works. All a court would require is evidence
that proves beyond a reasonable doubt that the work contains unlicensed
copyrighted material.
> > > Moreover, even if you could get a copyright in that combination, there
> > > is no problem with releaseing the kghostview/Qt combination under
> > > the GPL.
> >
> > 3b of the QPL prevents you from distributing QPLed code under the
> > terms of Section 2 of the GPL.
>
> Again, you don't understand what a copyright in a compilation or
> collective work is. The QPL has nothing to do with the copyright in
> the collective work (assuming there is such a thing in the kghostview
> example).
I did not claim that the QPL claims authority over the distribution terms
of the other pieces of the collective work. I claim that the GPL authority
over the distribution terms of the other pieces of the collective work,
and that the QPL conflicts with this claim of the GPL.
> > > Note that the collective work copyright does *not* reside in
> > > kghostview or in Qt; only in the combination (basically in the
> > > selection of which works to combine).
> >
> > You're now claiming that the author of kghostview just happened,
> > coincidentally, with no creativity, to come up with something that
> > works with Qt?
>
> No, but that has to do with the copyright on kghostview. You just
> don't get it.
Yes, exactly, the fact that the khghostview did not "just happen",
that it required creativity to create a work which includes libqt and
ghostview and others, is very much something that has to do with the
copyright on kghostview.
> > > I have no problem releasing my selection of combining kghostview and
> > > Qt under the GPL -- people are free to copy and change that selection
> > > however they please. What restrictions of the GPL would be violated?
> >
> > Section 6. [paraphrased: works based on the GPLed program must be
> > distributed under terms which are no more restrictive than the terms of
> > the GPL.]
>
> Right, and I can distribute my copyright in the kghostview/Qt
> combination under the GPL (I don't see where that is required, but
> assuming it were).
What do you mean by "distribute my copyright"? How is that significant
to this conversation.
kghostview -- especially working copies of kghostview -- is an example
of the "work based on the GPLed program", in this context.
> > Section 2b. [paraphrased: to publish kghostview, you must do so under
> > terms which don't allow Troll to relicense Qt mods such that the new
> > license would conflict with the GPL.]
>
> Again you are confused. You are talking about Qt source code again,
> whereas the discussion is about the (supposed) copyright in the
> collective work.
Let's see, we're talking about a collective work where the GPL is the
copyright on the collective work as a whole and the QPL is the copyright
on one of the pieces of that collective work.
You're currently trying to trivialize the creation of the collective
work as a whole, by claiming that distribution decisions -- made long
after the work as a whole was created -- are significant because they
do not involve enough creativity.
And, in this context, I'm confused because I can't figure out whether
we're talking about Qt or copyright on the collective work?
> > With each of these clauses, you do have to take other parts of the
> > license into account. However, there is nothing about section 6 or
> > about section 2b which lets you ignore any other parts of the license.
>
> The only person ignoring anything is you. You are ignoring the fact
> that the collective work copyright does not include the Qt copyright.
The collective work copyright is the GPL. Of course the GPL doesn't
include the QPL.
However, the collective work contains libqt, and the QPL applies to libqt.
And, the GPL requires that the libqt portion of the collective work,
as all other portions, be distributed under the terms of Section 2 of
the GPL. And, the QPL conflicts with those terms, preventing them from
being satisfied.
> > > (I note in this regard that the way the GPL is written it makes
> > > virtually no sense to apply it to a copyright in collective works).
> >
> > The author of the GPL would appear to disagree with you. Otherwise,
> > why would section 2 have the text:
> >
> > ... the intent is to exercise the right to control the distribution of
> > derivative or collective works based on the Program.
> >
> > ?
>
> Of course the GPL controls the distribution of collective works
> based on the Program. That does not mean, however, that it controls
> the licensing of the collective work itself; rather, it controls
> the licensing of the GPL component of the collective work and, to
> some degree, through the "no-charge" and "complete source code"
> requirements, the licensing of the other components of the collective
> work.
Ah, you're saying that the "under the terms of this license" clause have
no legal significance.
You're also saying that the wording of Section 6 has no legal
significance in the context of a collective work.
Right?
> > > Do not fall into the trap of thinking that the components of
> > > the collective work itself must be licensed the same way the
> > > collective work is.
> >
> > Sure, but in the case of the GPL, there's an explicit requirement
> > that the license on the components be no more restrictive than the
> > GPL itself.
>
> If it's that explicit, I wonder why there is no language quoted to
> support this statement. If you refer to Section 6, that only applies
> to the Program and works based on the Program, and Qt is neither.
This issue of yours is addressed by the top two paragraphs on the
third page of the GPL. It's very clear that if Qt is being distributed
by itself -- not as part of a collective work -- then there's no issue.
<quote>
But when you distribute the same sections as part of a whole which is
a work based on the Program, the distribution of the whole must be on
the terms of this License, whose permissions for other licensees extend
to the entire whole, and thus to each and every part regardless of who
wrote it.
</quote>
By the way, this is a part of the terms of Section 2.
> > And the GPL is very clear about what it means if the license on any of
> > the components is more restrictive than the GPL.
> >
> > > > I did not agree that X is not licensed under the GPL. I did agree
> > > > that X is licensed under XFree. I don't know why you have a problem
> > > > with this.
> > >
> > > For the reasons I stated in
> > > http://www.debian.org/Lists-Archives/debian-legal-0001/msg00151.html
> >
> > Example statement with false implications:
> >
> > ... the BSD license says you can modify the source code, it does not
> > permit you to modify the license. The copyright holder still holds
> > the copyright to every copy of the work that is created.
> >
> > With a collective work, there can be more than one copyright holder.
> > And, therefore, the copyrights which are granted are only the copyrights
> > granted by all copyright holders.
>
> I guess you mean you can only redistribute with permission of all relevant
> copyright holders.
Exactly.
And I also mean that the copyright license is the expression of that
permission.
> > If one copyright holder doesn't grant rights it doesn't matter that
> > another copyright holder does grant those rights..
> >
> > This is why BSD licensed material can be turned into proprietary,
> > commercial works, such as those advertised at www.f5.com.
>
> That can be done b/c BSD does not require code added to it to be
> licensed under the BSD; however, the BSD code still remains so
> licensed.
Yes, exactly.
The BSD code, considered by itself, is an independent work. However,
adequate permission has been granted to include it in a collective work
under any of a variety of collective licenses. And, the BSD license
doesn't require that people distributing the collective work make any
efforts to distribute the independent work represented by the BSD code
by itself.
> > > http://www.debian.org/Lists-Archives/debian-legal-0001/msg00160.html.
> >
> > Example statement with false implications:
> >
> > ... I suppose you can say that you are distributing the BSD code to me
> > both under BSD and GPL, and I must comply with both of them. Then I
> > wonder, but doesn't requiring me to comply with BSD (the advertising
> > restriction and the copyright notice) require me to violate the GPL?
> > Ahh, but you may say that is only a technical violation. ...
> >
> > No clauses from the BSD license were presented which conflicted with
> > any of the clauses from the GPL.
>
> The advertising clause is a "further restrictions on the recipients' exercise
> of the rights granted herein" (Section 6 of the GPL).
That clause was discarded years ago. The BSD license no longer contains it.
As I understand it, that clause was discarded because:
(1) it was unenforcable, and
(2) it conflicted with the GPL.
--
Raul
Reply to: