GPL Loophole or Not?

5 views
Skip to first unread message

Jimmy Cerra

unread,
Jul 22, 2004, 3:02:15 AM7/22/04
to
First of all, I'm not going to do this and think it would be a bad
thing to do. Now for the question:

I've read the General Public License, and one consequence is that a
software developer must release the any source code to a project
derived from another GPLed project at the (optional) cost of
distributing it (i.e. price of shipping). Well, what if a company:

1. Made a "patcher" program that took the source to a GPLed program.
2. Patched it.
3. Made an executable from the patched source.
4. Reverted the source back to its pre-patched state.

Now the patcher wasn't distributed under a license compatible with the
GPL. In fact, it was distributed under a draconian EULA. In this
hypothetical situation, was the GPL violated? Why or why not?

Possible reasons why the GPL was NOT violated:

* The company distributed the source to the origional work. The
closed source patcher program could have been distributed seperately.
You certainly have the right under the GPL to modify the source and
not give employees access to the source if they didn't distribute the
derived work to others.

* Compiling was done at the user's site, so the derived work (the
patched source) wasn't distributed.

A possible reason why the GPL was violated:

* The __intent__ was to distribute the patched program.

I am certainly stumped!

--
Jimmy Cerra

Alexander Terekhov

unread,
Jul 22, 2004, 4:51:03 AM7/22/04
to

Jimmy Cerra wrote:

[... patcher ...]

Portion(s) of the patcher program that contain "some" (non-"de
minimis" amount of) protected elements taken from the GPL'ed work
must be GPL'ed. If that is not the case (e.g. SUN's old converter
of Linux drivers to Solaris) then the GPL (or any other bare
copyright license) is totally irrelevant. Hth.

regards,
alexander.

Barry Margolin

unread,
Jul 22, 2004, 8:32:44 AM7/22/04
to
In article <9462edc9.04072...@posting.google.com>,
jimb...@hotmail.com (Jimmy Cerra) wrote:

> First of all, I'm not going to do this and think it would be a bad
> thing to do. Now for the question:
>
> I've read the General Public License, and one consequence is that a
> software developer must release the any source code to a project
> derived from another GPLed project at the (optional) cost of
> distributing it (i.e. price of shipping). Well, what if a company:
>
> 1. Made a "patcher" program that took the source to a GPLed program.
> 2. Patched it.
> 3. Made an executable from the patched source.
> 4. Reverted the source back to its pre-patched state.
>
> Now the patcher wasn't distributed under a license compatible with the
> GPL. In fact, it was distributed under a draconian EULA. In this
> hypothetical situation, was the GPL violated? Why or why not?

I don't know whether it would hold up in court, but I believe that the
FSF would consider it a violation. I base this on the fact that it's
similar to another situation that they claimed was a violation. This
was a program that needed to be linked with the GMP library. The author
distributed this program as object files, along with instructions on how
to download the GMP library and link it. The FSF considered this
effectively equivalent to distributing the result of the linking -- the
fact that the linking takes place at the recipient's site is a minor
detail that shouldn't affect the legal status.

Your scenario replaces the instructions on how to link the program with
a patcher program, but it's still pretty much the same thing.

--
Barry Margolin, bar...@alum.mit.edu
Arlington, MA

Alexander Terekhov

unread,
Jul 22, 2004, 9:05:26 AM7/22/04
to

Barry Margolin wrote:
[...]

> FSF would consider it a violation. I base this on the fact that it's
> similar to another situation that they claimed was a violation. This
> was a program that needed to be linked with the GMP library. The author
> distributed this program as object files, along with instructions on how
> to download the GMP library and link it. The FSF considered this
> effectively equivalent to distributing the result of the linking -- the
> fact that the linking takes place at the recipient's site is a minor
> detail that shouldn't affect the legal status.

Right, it doesn't affect the legal status. The court would consider
it a "violation" of doctrine of copyright misuse (good-bye all FSF's
copyrights) and first sale. The only reasonable conclusion would be
"Hey FSF, stuff it and pay the defendant's legal expenses". We can
call the movie "SCO reloaded" (or something like that).

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 22, 2004, 12:00:40 PM7/22/04
to
On Thu, 22 Jul 2004 15:05:26 +0200
Alexander Terekhov <tere...@web.de> wrote:

> The court would consider it a "violation" of doctrine
> of copyright misuse (good-bye all FSF's copyrights)

Care to explain why the outcome of a single court
case would have an influence on all the copyrights
held by the FSF?

--
Stefaan
--
"What is stated clearly conceives easily." -- Inspired sales droid

Alexander Terekhov

unread,
Jul 22, 2004, 1:20:09 PM7/22/04
to

Stefaan A Eeckels wrote:
>
> On Thu, 22 Jul 2004 15:05:26 +0200
> Alexander Terekhov <tere...@web.de> wrote:
>
> > The court would consider it a "violation" of doctrine
> > of copyright misuse (good-bye all FSF's copyrights)
>
> Care to explain why the outcome of a single court
> case would have an influence on all the copyrights
> held by the FSF?

Because the penalty for copyright misuse is severe copyright
impotence (Viagra resistant). FSF's GPL FAQ absurdity "covers"
all FSF's copyrights, for example.

regards,
alexander.

Barry Margolin

unread,
Jul 22, 2004, 2:20:51 PM7/22/04
to
In article <40FFBB96...@web.de>,
Alexander Terekhov <tere...@web.de> wrote:

I'm confused. Your first sentence suggests that you agree with me, but
then you come to the exact opposite conclusion. Are you claiming that
the GPL would not apply when the redistributor performs the linking
himself before redistributing copies?

Alexander Terekhov

unread,
Jul 22, 2004, 2:27:38 PM7/22/04
to

Barry Margolin wrote:
[...]

> I'm confused. Your first sentence suggests that you agree with me, but
> then you come to the exact opposite conclusion. Are you claiming that
> the GPL would not apply when the redistributor performs the linking
> himself before redistributing copies?

Forget linking. Linking is irrelevant. First sale aside for a
moment, redistribution of unmodified copies is governed by the
GPL section 1 and section 3, not section 2.

regards,
alexander.

Barry Margolin

unread,
Jul 22, 2004, 5:04:50 PM7/22/04
to
In article <4100071A...@web.de>,
Alexander Terekhov <tere...@web.de> wrote:

But linking produces modified copies, not unmodified copies, so why is
it irrelevant?

And the OP's scenario involved a patch script, which obviously creates
modified copies. If the patch script were run before redistribution,
you'd clearly be redistributing modified copies. Should someone really
be able to make an end-run around the license by having the end user run
the patch script? Is it sensible that two processes that have the same
end result should be treated so differently, based only on whether the
patch script is run before or after redistribution? (Yes, I realize
that the law is not always logical, but I harbor some small hope that
judges and juries will apply common sense every now and then.)

Alexander Terekhov

unread,
Jul 22, 2004, 5:45:50 PM7/22/04
to

Barry Margolin wrote:
>
> In article <4100071A...@web.de>,
> Alexander Terekhov <tere...@web.de> wrote:
>
> > Barry Margolin wrote:
> > [...]
> > > I'm confused. Your first sentence suggests that you agree with me, but
> > > then you come to the exact opposite conclusion. Are you claiming that
> > > the GPL would not apply when the redistributor performs the linking
> > > himself before redistributing copies?
> >
> > Forget linking. Linking is irrelevant. First sale aside for a
> > moment, redistribution of unmodified copies is governed by the
> > GPL section 1 and section 3, not section 2.
>
> But linking produces modified copies, not unmodified copies, so why is

http://groups.google.com/groups?selm=40F3AA16.315F565D%40web.de

> it irrelevant?
>
> And the OP's scenario involved a patch script, which obviously creates
> modified copies.

First bring to justice http://babelfish.altavista.com/ -- we can talk
then.

regards,
alexander.

Lee Hollaar

unread,
Jul 22, 2004, 6:10:56 PM7/22/04
to
In article <barmar-669C02....@comcast.dca.giganews.com> Barry Margolin <bar...@alum.mit.edu> writes:
> Is it sensible that two processes that have the same
>end result should be treated so differently, based only on whether the
>patch script is run before or after redistribution?

Sure, if a license agreement allows one but does not allow the other.

The writers of a license (especially one that is a take-it-or-leave
it one like the GPL, where no changes to the terms are allowed) are
in the best position to frame the terms of the license in the way
that gives them the protection and control that they want. A general
priciple of contract law is that disputed terms are construed against
the one writing the terms, since he was in the best position to get
the terms the way he wanted.

The drafters of the GPL, for whatever reason, decided not to impose
terms regarding the ownership or use of the licensed software.
Instead, the terms have to do with the redistribution.

Since there is no claim made that the user does not own his or her
particular copy of the licensed software (as opposed to most
proprietary software, where the license claims that the company
retains ownership), in the United States a user gets to make whatever
copies are necessary to use the program and any adaptations that
are needed. See 17 USC 117.

But Section 117 does not allow the distribution of the adaptations
without the permission of the copyright owner. And that alone
provides a good reason why there is a difference in result between
patching and redistributing, and user patching. They are treated
differently under copyright law.

The lesson is that if you want to prohibit something, you frame your
license to prohibit it, which may involve tricks like saying that
the user doesn't own the software to avoid the user rights of
Section 117.

Isaac

unread,
Jul 22, 2004, 9:43:52 PM7/22/04
to

It would not be an end run. The GPL deliberately allows users to do
things that copyright law says that the author does not have to allow.
No one seems to dispute that users can create the derivative work
on their own platform. It also does not seem to be contested that
it's okay for users to contract for others to create derivative works
on the user's computer.

So given that, why does it make a difference whether I come to your
place and create the derivative work (and then delete all of the
sources) or if I just email you the thing charging you exactly the same
amount of money.

I don't find argument that the two processes with the same end result
must be treated the same under the law to be persuasive. It is easy
to find things in the law where the process of accomplishing something
does matter.

Isaac

Isaac

unread,
Jul 22, 2004, 9:47:10 PM7/22/04
to
On Thu, 22 Jul 2004 19:20:09 +0200, Alexander Terekhov <tere...@web.de> wrote:
>
>
> Because the penalty for copyright misuse is severe copyright
> impotence (Viagra resistant). FSF's GPL FAQ absurdity "covers"
> all FSF's copyrights, for example.
>

I've probably missed the posts where you have explained this, but
what is your description of what the law considers copyright abuse?
What I've read about the topic does not seem applicable to the FSF.

Isaac

Stefaan A Eeckels

unread,
Jul 22, 2004, 5:51:09 PM7/22/04
to
On Thu, 22 Jul 2004 19:20:09 +0200
Alexander Terekhov <tere...@web.de> wrote:

> Stefaan A Eeckels wrote:
> >
> > On Thu, 22 Jul 2004 15:05:26 +0200
> > Alexander Terekhov <tere...@web.de> wrote:
> >
> > > The court would consider it a "violation" of doctrine
> > > of copyright misuse (good-bye all FSF's copyrights)
> >
> > Care to explain why the outcome of a single court
> > case would have an influence on all the copyrights
> > held by the FSF?
>
> Because the penalty for copyright misuse is severe copyright
> impotence (Viagra resistant). FSF's GPL FAQ absurdity "covers"
> all FSF's copyrights, for example.

You're usually so URL-happy I'm amazed you didn't
include a reference to a law text or judgement.

Seriously though, are you suggesting that when a
plaintiff would lose a copyright infringment case,
_all_ their works could be declared in the public
domain?

Lee Hollaar

unread,
Jul 22, 2004, 10:27:05 PM7/22/04
to
In article <20040722235109...@DELETEMEecc.lu> Stefaan A Eeckels <te...@DELETEMEecc.lu> writes:
>On Thu, 22 Jul 2004 19:20:09 +0200
>Alexander Terekhov <tere...@web.de> wrote:
>
>> Stefaan A Eeckels wrote:
>> >
>> > On Thu, 22 Jul 2004 15:05:26 +0200
>> > Alexander Terekhov <tere...@web.de> wrote:
>> >
>> > > The court would consider it a "violation" of doctrine
>> > > of copyright misuse (good-bye all FSF's copyrights)
>> >
>> > Care to explain why the outcome of a single court
>> > case would have an influence on all the copyrights
>> > held by the FSF?
>>
>> Because the penalty for copyright misuse is severe copyright
>> impotence (Viagra resistant). FSF's GPL FAQ absurdity "covers"
>> all FSF's copyrights, for example.
>
>You're usually so URL-happy I'm amazed you didn't
>include a reference to a law text or judgement.

Here's a couple --

First, a discussion of copyright misuse from my treatise:
http://digital-law-online.info/lpdi1.0/treatise15.html#secII.K.

Then, the leading case on the subject -- Lasercomb v. Reynolds:
http://digital-law-online.info/lpdi1.0/cases/15PQ2D1846.htm

>Seriously though, are you suggesting that when a
>plaintiff would lose a copyright infringment case,
>_all_ their works could be declared in the public
>domain?

Copyright misuse is a strange thing. You don't lose your copyright,
but you lose your right to enforce it in court until you completely
purge your misuse.

So, for example, if the GPL were to be found to be a misuse of
copyright, because it improperly tries to extend the protection
beyond that intended by Congress (such as, for example, saying
that you couldn't prepare a competitive work for 99 years), then
the GPL and the copyrights underlying it could become unenforceable.

But it's uncommon that copyright misuse is found, and certainly it's
not something that happens just because you lose a copyright suit.

Alexander Terekhov

unread,
Jul 23, 2004, 4:56:44 AM7/23/04
to

Isaac wrote:
>
> On Thu, 22 Jul 2004 19:20:09 +0200, Alexander Terekhov <tere...@web.de> wrote:
> >
> >
> > Because the penalty for copyright misuse is severe copyright
> > impotence (Viagra resistant). FSF's GPL FAQ absurdity "covers"
> > all FSF's copyrights, for example.
> >
>
> I've probably missed the posts where you have explained this, but
> what is your description of what the law considers copyright abuse?

I've posted numerous times (on gnu.misc.discuss) links to the
messages of mine pointing folks (deceptively bluffed by the FSF's
hypocrites) to Hollar's treatise and the devastating legal review
of the GPL written by Christian H. Nadan (the author of "A Proposal
to Recognize Component Works: How a Teddy Bears on the Competing
Ends of Copyright Law" cited in LEWIS GALOOB TOYS, INC. v. NINTENDO
OF AMERICA, INC.).

http://groups.google.com/groups?selm=40EE745B.7A2ED0DF%40web.de
http://groups.google.com/groups?selm=40EE8CDC.977AE902%40web.de

Try also this:

http://tinyurl.com/249sg
http://tinyurl.com/3jg8z

Hth.

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 23, 2004, 4:59:37 AM7/23/04
to
On Fri, 23 Jul 2004 02:27:05 +0000 (UTC)
hol...@faith.cs.utah.edu (Lee Hollaar) wrote:

> In article <20040722235109...@DELETEMEecc.lu> Stefaan A
> Eeckels <te...@DELETEMEecc.lu> writes:
> >On Thu, 22 Jul 2004 19:20:09 +0200
> >Alexander Terekhov <tere...@web.de> wrote:
> >>
> >> Because the penalty for copyright misuse is severe copyright
> >> impotence (Viagra resistant). FSF's GPL FAQ absurdity "covers"
> >> all FSF's copyrights, for example.
> >
> >You're usually so URL-happy I'm amazed you didn't
> >include a reference to a law text or judgement.
>
> Here's a couple --
>
> First, a discussion of copyright misuse from my treatise:
> http://digital-law-online.info/lpdi1.0/treatise15.html#secII.K.
>
> Then, the leading case on the subject -- Lasercomb v. Reynolds:
> http://digital-law-online.info/lpdi1.0/cases/15PQ2D1846.htm

Thanks for the references.

> >Seriously though, are you suggesting that when a
> >plaintiff would lose a copyright infringment case,
> >_all_ their works could be declared in the public
> >domain?
>
> Copyright misuse is a strange thing. You don't lose your copyright,
> but you lose your right to enforce it in court until you completely
> purge your misuse.

OK, that makes sense. It would still not be OK to treat
such work as public domain, because one could be sued once
the misuse has been purged. A prudent operator will refrain
from engaging in actions that illegal because the other
party did something illegal, at least in my book.

> So, for example, if the GPL were to be found to be a misuse of
> copyright, because it improperly tries to extend the protection
> beyond that intended by Congress (such as, for example, saying
> that you couldn't prepare a competitive work for 99 years), then
> the GPL and the copyrights underlying it could become unenforceable.

The GPL is careful in not asserting rights to the
code that is not a derivative work of the GPLed work.
It assumes (like I do) that a binary executable program
is a derivative work of all its composing source codes,
hence covered by the copyrights of all these works.

> But it's uncommon that copyright misuse is found, and certainly it's
> not something that happens just because you lose a copyright suit.

From perusing your references, I understand that misuse
would be conditions present in the license, i.e. part of the
understanding between the parties. Hence a FAQ document
explaining the copyright holder's interpretation of a
"derivative work", not being part of the license, would
not lead to a conclusion of copyright misuse. A case would
be judged on the merits of the license, not the interpretation
of either party of the meaning of the statutes (as it is
essentially the rôle of the court to settle these types
of differences).

Take care,

Alexander Terekhov

unread,
Jul 23, 2004, 6:31:35 AM7/23/04
to

Stefaan A Eeckels wrote:
[...]

> It assumes (like I do) that a binary executable program
> is a derivative work of all its composing source codes,
> hence covered by the copyrights of all these works.

Executable isn't a derivative work, to begin with. As for the rest,

http://groups.google.com/groups?selm=40FCE432.4F633F87%40web.de
(gpl.txt -> LGPL)

http://groups.google.com/groups?selm=40EDB041.E681930C%40web.de
(LGPL definitions and FSF's intent... conduct/impact aside for a moment)

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 23, 2004, 8:14:35 AM7/23/04
to
On Fri, 23 Jul 2004 12:31:35 +0200
Alexander Terekhov <tere...@web.de> wrote:

> Stefaan A Eeckels wrote:
> [...]
> > It assumes (like I do) that a binary executable program
> > is a derivative work of all its composing source codes,
> > hence covered by the copyrights of all these works.
>
> Executable isn't a derivative work, to begin with.

Jurisprudence please.

Alexander Terekhov

unread,
Jul 23, 2004, 8:31:41 AM7/23/04
to

Stefaan A Eeckels wrote:
>
> On Fri, 23 Jul 2004 12:31:35 +0200
> Alexander Terekhov <tere...@web.de> wrote:
>
> > Stefaan A Eeckels wrote:
> > [...]
> > > It assumes (like I do) that a binary executable program
> > > is a derivative work of all its composing source codes,
> > > hence covered by the copyrights of all these works.
> >
> > Executable isn't a derivative work, to begin with.
>
> Jurisprudence please.

Derivative work ("Bearbeitung") must contain new protected
expressions in order to be a derivative work ("Umgestaltungen"
aside for a moment). Executable is just another form of a
corresponding source tarball (a compilation of sources and
whatever resources). It's an aggregation of copies in binary
form. Try to imagine a computer system that can run tarballs
directly. See the light?

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 23, 2004, 9:27:42 AM7/23/04
to
On Fri, 23 Jul 2004 14:31:41 +0200

One could argue that by combining existing source codes with
new source codes one does something like writing a sequel,
mutatis mutandis. IIRC, merely selecting information for
inclusion in a new work can be sufficient to enjoy copyright
protection.

The problem is that analogies are never a perfect match.
Obviously, you feel that using existing source code in a
new program is merely placing them together into a kind of
archive. To me, it feels like copying paragraphs from
a text - plagiarism. A little while ago, I read a document
that had been "written" through cutting and pasting sections
and paragraphs from other documents (written mainly by yours
truly), sections culled from Sun's Java site, and some
new material. All the existing documents could be freely
downloaded. Using your approach, it would be the same as
buying a couple of pocketbooks and making a collage.
I submit that it was plain, unadulterated plagiarism
(especially because said document was devoid of any
attribution), and hence a copyright violation (in addition
to being thoroughly unprofessional). I can see little
if any difference between this (ab)use of copyrighted material
and the use of GPLed source code to create a new program.

Take care,

Alexander Terekhov

unread,
Jul 23, 2004, 10:05:04 AM7/23/04
to

Stefaan A Eeckels wrote:
[...]
> a text - plagiarism. A little while ago, I read a document
> that had been "written" through cutting and pasting sections
> and paragraphs from other documents (written mainly by yours
> truly), sections culled from Sun's Java site, and some
> new material. All the existing documents could be freely
> downloaded. Using your approach, it would be the same as
> buying a couple of pocketbooks and making a collage.
> I submit that it was plain, unadulterated plagiarism
> (especially because said document was devoid of any
> attribution),

That's bad. Attribution/acknowledgment (moral right to claim
authorship) is established by the Berne Convention and is
already the mandatory requirement in some jurisdictions (e.g.
Australia... since 2000, IIRC).

> and hence a copyright violation (in addition
> to being thoroughly unprofessional).

Funny. Then explain to me why the FSF decided to proclaim
Xfree86 1.1 license that now includes so-called acknowledgment
clause

--
3. The end-user documentation included with the redistribution,
if any, must include the following acknowledgment: "This product
includes software developed by The XFree86 Project, Inc
(http://www.xfree86.org/) and its contributors", in the same
place and form as other third-party acknowledgments. Alternately,
this acknowledgment may appear in the software itself, in the
same form and location as other such third-party acknowledgments.
--

"GPL incompatible" (whatever that means)?

> I can see little
> if any difference between this (ab)use of copyrighted material
> and the use of GPLed source code to create a new program.

It's not a "new program". Its a combination of GPL'ed and non-
GPL'ed "programs" (literary works that are functional but are
still protected as literary works, not as functional artifacts)
working together. They are separate works, not derivatives of
each other. FSF concedes that things like IPC, sockets, and
CORBA, establish a boundary that the GPL doesn't cross... well,
but linking "intimacy" is completely beyond the scope of the
copyright law.

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 23, 2004, 11:03:44 AM7/23/04
to
On Fri, 23 Jul 2004 16:05:04 +0200
Alexander Terekhov <tere...@web.de> wrote:

> It's not a "new program".

Sorry to be long-winded but you have a knack for evading
the question. Could you please try to indicate what you
consider the impact of using GPLed source code in a new
development?

If one produces a program that contains new source code
as well as source code taken from GPLed work, and compiles
it into a single unit that contains binary code corresponding
to both the new source code and the GPLed source code, do
you believe that
a) The executable can be distributed without any
source code
b) The executable can be distributed with only the
source code taken from the GPLed program
c) The executable can only be distributed with the
complete source code under the GPL?

We're not arguing the relative merits, contributions,
or volume of the components, and assume that the
volume of the GPLed source code is large enough to
be outside of a "fair use" claim.

If you choose a) or b), please explain how this differs
from preparing a new document by cutting and pasting
sections and paragraphs from existing documents, with or
without attribution, and again, outside of any fair use
claims.

Thanks.

> Its a combination of GPL'ed and non-
> GPL'ed "programs" (literary works that are functional but are
> still protected as literary works, not as functional artifacts)
> working together. They are separate works, not derivatives of
> each other.

I have problems considering a single executable "separate
works" (static linking). The case I was describing is the
use of source code lifted from a GPLed work to create a new,
functionally different work. The new work contains copyrightable
material from the GPLed work, as well as material written
by the programmer using source from the GPLed work.

I am _not_ arguing that dynamic linking is the same as
static linking, neither am I arguing that the use of a
library makes a program a derivative work of that
library (or that the use of OS system calls makes a
program a derivative work of the OS).

What I _do_ argue is that independently of the availability
of the source code, the use of any source code in a new work
is subject to copyright, and not a case of first sale.
To state it unambiguously, if I take functions from a GPLed work
in source format, add them to my program source directory,
and compile the lot into an executable, I do something that
is the same as cutting and pasting sections from a document
into a new document. I also argue that extracting the compiled
versions of the sources from an existing library and producing
an executable that contains this compiled code is the same
as copying of the source.

In other words, if I publish a piece of software under the GPL,
you cannot download a copy, extract the source files you like,
and use them in a program you're writing without honouring the
license because the resulting executable is a "collage" of
works obtained under first sale and your own work. It doesn't
matter that you can physically download the GPLed work as often
as you please.

> FSF concedes that things like IPC, sockets, and
> CORBA, establish a boundary that the GPL doesn't cross... well,
> but linking "intimacy" is completely beyond the scope of the
> copyright law.

Indeed. That is why IMVHO a dynamically linked executable is
not a derivative work of the libraries it calls, whereas a
program that physically contains copyrighted materials is
(if not a derivative work) subject to the copyrights and
licenses of all its components.

Alexander Terekhov

unread,
Jul 23, 2004, 11:41:18 AM7/23/04
to

Stefaan A Eeckels wrote:
[...]
> If you choose a) or b),

I choose b) with proper attribution/acknowledgment, of course.

http://groups.google.com/groups?selm=40276AEC.EF2241DF%40web.de
(Subject: Re: GPL warning label idea, has anyone thought of this before?)

> please explain how ...

http://groups.google.com/groups?selm=40F29FA5.EECEC5E6%40web.de
(Subject: Re: Use of GPL'd code with proprietary programs)

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 23, 2004, 12:42:32 PM7/23/04
to
On Fri, 23 Jul 2004 17:41:18 +0200
Alexander Terekhov <tere...@web.de> wrote:

> Stefaan A Eeckels wrote:
> [...]
> > If you choose a) or b),
>
> I choose b) with proper attribution/acknowledgment, of course.

If you were to use source code that is available on
a web site, but not under the GPL or any other license
stipulating terms of use, would a resulting executable


that contains binary code corresponding to both the new

source code and the non-licensed source code be

a) distributable without any source code and no
attributions
b) distributable without any source code and
proper attributions
c) not distributable

It would seem to me that the only possible answer is c),
you would _not_ be allowed to use the source in a program
you write, just as you're not allowed to use parts of a
document written by someone else in your own documents
("fair use" is pretty limited and is not merely determined
by the size of the copied material), with or without proper
attribution.

> http://groups.google.com/groups?selm=40276AEC.EF2241DF%40web.de
> (Subject: Re: GPL warning label idea, has anyone thought of this
> before?)
>
> > please explain how ...
>
> http://groups.google.com/groups?selm=40F29FA5.EECEC5E6%40web.de
> (Subject: Re: Use of GPL'd code with proprietary programs)

In this reference you say:

| If you don't create "a work based upon one or more preexisting works"
| then you have simply not created a derivative work. 17 U.S.C. §101.

It would seem fairly obvious that by taking source code from
a GPLed work you do "base your work upon one or more preexisting
works", don't you think so? How is this different from creating
a document by taking paragraphs from a preexisting document?

We're not talking about two independently written pieces of
software. It's a single piece of software that contains source
code you wrote, and source code someone else wrote.

Alexander Terekhov

unread,
Jul 23, 2004, 1:06:25 PM7/23/04
to

Stefaan A Eeckels wrote:
[...]
> a) distributable without any source code and no
> attributions
> b) distributable without any source code and
> proper attributions
> c) not distributable

I'll assume that a permission is given to "take it" (not merely
read or study it as some material merely presented on someone's
web site, in a book, or whatever).

Of course b). I believe that the GPL condition to make the source
code available is preempted by the first sale (as long as I don't
modify it). I just see nothing wrong with passing the GPL'ed
source code along with my proprietary/whatever stuff. Take it and
do whatever you want.

[...]


> We're not talking about two independently written pieces of
> software. It's a single piece of software that contains source
> code you wrote, and source code someone else wrote.

It not a "single piece". All I had was a functional spec for the
GPL'ed program/code. It's a back box as far as I'm concerned. The
resulting aggregation is "packaged" as a single executable, but
that doesn't change anything. It's the same (as far as copyright
is concerned) as a bunch of DLLes put together with "only small
main function" executable.

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 23, 2004, 7:05:53 PM7/23/04
to
On Fri, 23 Jul 2004 19:06:25 +0200
Alexander Terekhov <tere...@web.de> wrote:

>
> Stefaan A Eeckels wrote:
> [...]
> > a) distributable without any source code and no
> > attributions
> > b) distributable without any source code and
> > proper attributions
> > c) not distributable
>
> I'll assume that a permission is given to "take it" (not merely
> read or study it as some material merely presented on someone's
> web site, in a book, or whatever).

Like "djbdns" on Dan Bernstein's site, for example.

> Of course b). I believe that the GPL condition to make the source
> code available is preempted by the first sale (as long as I don't
> modify it). I just see nothing wrong with passing the GPL'ed
> source code along with my proprietary/whatever stuff. Take it and
> do whatever you want.

I would feel perfectly happy using code I legally obtained
for my own needs, but using it as part of a development
I want to distribute (for free or for money) is not on.
Once it comes to incorporating someone else's source code
I believe copyright has the same implications for source
code as it has for written documents.
Note that my question was not about the GPL, but about source
code you obtained legally - like a book you obtained legally,
the only difference being you got it from a web site.
You effectively say that once you obtained one copy, no amount
of copying the files around on your own system(s), integrating
them in executables in compiled format, and copying those
executables for distribution is prohibited by the copyright
statutes, because of "first sale". At least, that's how I
understand "do whatever you want".

> [...]
> > We're not talking about two independently written pieces of
> > software. It's a single piece of software that contains source
> > code you wrote, and source code someone else wrote.
>
> It not a "single piece". All I had was a functional spec for the
> GPL'ed program/code. It's a back box as far as I'm concerned. The
> resulting aggregation is "packaged" as a single executable, but
> that doesn't change anything. It's the same (as far as copyright
> is concerned) as a bunch of DLLes put together with "only small
> main function" executable.

Again, you're evading the question. Do you assert that using
source code (functions, subroutines, classes) written by
someone else as part of your development is not a copyright
violation? Then, obviously, neither is copying text written
by someone else into documents you write.

Once you present the GPLed software as a separate program,
I agree that you can place it alongside one of your own
creations, and the license of the one doesn't influence the
license of the other. But this is not what I asked.
I specifically asked if you think that lifting source code
from a legally obtained copy, and inserting it into your
own development is different from copying a section from
a document in your own writings.

I don't believe it is. I don't believe that availability
of multiple copies gives you the possibility to do what
has always been seen as unlawfully using someone else's
work simply by claiming first sale.

Alexander Terekhov

unread,
Jul 24, 2004, 1:24:15 PM7/24/04
to

Stefaan A Eeckels wrote:
[...]
> I would feel perfectly happy using code I legally obtained
> for my own needs, but using it as part of a development

But if you own that copy, you can sell it (with or without some add-on
stuff). That's what the first sale is about. Dissemination of property
(copyrighted works).

regards,
alexander.

Barry Margolin

unread,
Jul 24, 2004, 1:44:55 PM7/24/04
to
In article <41029B3F...@web.de>,
Alexander Terekhov <tere...@web.de> wrote:

I believe Alexander is operating under the belief that he can apply the
first-sale doctrine because instead of making new copies when he
redistributes, he can distribute the copy he downloaded, and then
download another one legally. So each time he goes through this, he's
simply disposing of the copy that he owns rather than doing anything
that the copyright owner has control over.

Alexander Terekhov

unread,
Jul 24, 2004, 2:17:06 PM7/24/04
to

Barry Margolin wrote:
[...]

> I believe Alexander is operating under the belief that he can apply the
> first-sale doctrine because instead of making new copies when he
> redistributes, he can distribute the copy he downloaded, and then
> download another one legally. So each time he goes through this, he's
> simply disposing of the copy that he owns rather than doing anything
> that the copyright owner has control over.

Forget downloading for a moment. The GPL does allow verbatim copying.
That's sections 1 and 3 (for object code form), not section 2. The GPL
grants the right to make copies and I own the copies I made. The only
way a court can interpret that license is to determine whether what is
being made -- A+B, for example, where A is the GPL-licensed work --
involves making a copy of A or creating a derivative work of A. If
the former, then the license is clear that there are no reciprocal
obligations. If the latter, then the license is also clear that the
author of "A+B" must disclose his source code. (these are most not
my words. Attribution: http://rosenlaw.com/html/lawrenceerosen.html)

regards,
alexander.

Barry Margolin

unread,
Jul 24, 2004, 3:08:41 PM7/24/04
to
In article <4102A7A2...@web.de>,
Alexander Terekhov <tere...@web.de> wrote:

> Forget downloading for a moment. The GPL does allow verbatim copying.
> That's sections 1 and 3 (for object code form), not section 2. The GPL
> grants the right to make copies and I own the copies I made.

I think that's one of the things most of us disagree with. You don't
"own" anything when you make a copy, except maybe for the disk drive you
put the copy on. For the first-sale doctrine to apply, you have to have
engaged in a transaction where a tangible object (e.g. a book, a
painting, CD, floppy disk) was transferred to you. The first-sale
doctrine applies to disposition of that object (so you can sell a book
to a used book store, or make a collage by cutting up magazines), *not*
the intellectual property. The first-sale doctrine makes a distinction
between the physical container and the information it contains.

When only bits are involved, how can you distinguish an "original" (to
which the first-sale doctrine applies) from a "copy" (to which it
doesn't)? The seems like it should make it impossible for the doctrine
to be used.

Alexander Terekhov

unread,
Jul 24, 2004, 3:24:43 PM7/24/04
to

Barry Margolin wrote:
[...]

> I think that's one of the things most of us disagree with. You don't
> "own" anything when you make a copy, except maybe for the disk drive you
> put the copy on.

Or the CD that contains the entire distribution. If you insist, I can
refrain from media-less distribution, no problem with that.

http://groups.google.com/groups?selm=40D46A0A.76407979%40web.de
(Subject: Re: The worst that can happen to GPLed code)

regards,
alexander.

Lee Hollaar

unread,
Jul 24, 2004, 3:59:36 PM7/24/04
to
In article Barry Margolin <bar...@alum.mit.edu> writes:
>I think that's one of the things most of us disagree with. You don't
>"own" anything when you make a copy, except maybe for the disk drive you
>put the copy on.

In United States copyright law, you don't "put the copy" on anything.
You put a "work" on some tangible medium, giving you a "copy"
"Copies" are material objects, other than phonorecords, in which a
work is fixed by any method now known or later developed, and from
which the work can be perceived, reproduced, or otherwise communicated,
either directly or with the aid of a machine or device. The term
"copies" includes the material object, other than a phonorecord, in
which the work is first fixed.
17 USC 101.

> For the first-sale doctrine to apply, you have to have
>engaged in a transaction where a tangible object (e.g. a book, a
>painting, CD, floppy disk) was transferred to you.

Nope. You only have to be "the owner of a particular copy or phonorecord
lawfully made" under copyright law. 17 USC 109. A copy can be "lawfully
made" if it is made by the copyright owner, made with the authorization
of the copyright owner, or made under one of the exceptions to the copyright
owner's exclusive rights.

> The first-sale
>doctrine applies to disposition of that object (so you can sell a book
>to a used book store, or make a collage by cutting up magazines), *not*
>the intellectual property. The first-sale doctrine makes a distinction
>between the physical container and the information it contains.

First sale looks to whether you have ownership of a material thing ("copy")
that was lawfully made. It does not look to whether you have ownership
of the copyrighted material ("work"), because the owner of a copyright
doesn't need first sale in order to transfer ownership to another.


>When only bits are involved, how can you distinguish an "original" (to
>which the first-sale doctrine applies) from a "copy" (to which it
>doesn't)? The seems like it should make it impossible for the doctrine
>to be used.

There are no "originals" in United States copyright law, only "copies."
See above: "The term 'copies' includes the material object ... in
which the work is first fixed."

The test for first sale is not whether you own an original or a copy,
but whether the copy you own was lawfully made.


Barry Margolin

unread,
Jul 24, 2004, 4:00:18 PM7/24/04
to
In article <4102B77B...@web.de>,
Alexander Terekhov <tere...@web.de> wrote:

> Barry Margolin wrote:
> [...]
> > I think that's one of the things most of us disagree with. You don't
> > "own" anything when you make a copy, except maybe for the disk drive you
> > put the copy on.
>
> Or the CD that contains the entire distribution. If you insist, I can
> refrain from media-less distribution, no problem with that.

But for first-sale to be applicable, you would have to transfer the CD
that you received, and remove any copies that you installed onto your
computer.

Alexander Terekhov

unread,
Jul 24, 2004, 4:10:02 PM7/24/04
to

Barry Margolin wrote:
[...]
> > > I think that's one of the things most of us disagree with. You don't
> > > "own" anything when you make a copy, except maybe for the disk drive you
> > > put the copy on.
> >
> > Or the CD that contains the entire distribution. If you insist, I can
> > refrain from media-less distribution, no problem with that.
>
> But for first-sale to be applicable, you would have to transfer the CD

No. I just need to prove that copies that reside on the CD were
lawfully made. Making them in the course of authorized downloading
is lawful, for example. And I don't think that the law forbids me
to store multiple works in an archive/tar (or some "executable")
on a CD (or whatever "tangible") and sell it.

regards,
alexander.

Lee Hollaar

unread,
Jul 24, 2004, 4:23:05 PM7/24/04
to
In article <4102C21A...@web.de> Alexander Terekhov <tere...@web.de> writes:
> I just need to prove that copies that reside on the CD were
>lawfully made.

This will be a more meaningful discussion if we get the language
right, since it's important for first sale and other copyright
principles.

"Copies" are material objects. The CD is a "copy."

"Works" are the things fixed in a "copy." There can be many
"works" fixed on a CD, each being a "copy."

Alexander Terekhov

unread,
Jul 24, 2004, 4:31:05 PM7/24/04
to

Ok. I just need to prove that all works fixed on the CD are lawfully
made copies. Right?

regards,
alexander.

Isaac

unread,
Jul 24, 2004, 4:33:56 PM7/24/04
to
On Sat, 24 Jul 2004 15:08:41 -0400, Barry Margolin <bar...@alum.mit.edu> wrote:
> In article <4102A7A2...@web.de>,
> Alexander Terekhov <tere...@web.de> wrote:
>
>> Forget downloading for a moment. The GPL does allow verbatim copying.
>> That's sections 1 and 3 (for object code form), not section 2. The GPL
>> grants the right to make copies and I own the copies I made.
>
> I think that's one of the things most of us disagree with. You don't
> "own" anything when you make a copy, except maybe for the disk drive you
> put the copy on. For the first-sale doctrine to apply, you have to have
> engaged in a transaction where a tangible object (e.g. a book, a
> painting, CD, floppy disk) was transferred to you. The first-sale

While I don't subscribe to Terekhov's first sale theory, I don't think
there is any question that I own the any copies of software I legally
make on my own system. Further, first-sale in the US is not merely
doctrine, it is statutory. I don't see any words in the statute requiring
transfer of a tangible object. I would expect first sale to apply to
software that I paid for and received by downloading from the Internet
as long as I was willing to delete my own copy when transfering it to
another (ignoring for now the question of whether I owned rather than
licensed the software).

There may be a loophole in the literal wording of the statute, but I
don't believe a court will necessarily interpret the words the way
Terehkov suggests if it believes that doing so is contrary to the
intent of Congress.

IMO the statute (17 USC 109) does contain a buglet that could allow
distribution of fair use copies, computer backups, and other stuff,
against the will of the copyright holder. I don't believe a court
would allow the bug to be exploited. I believe instead that the
court will consider what was done with the copies when deciding
whether the copies were lawfully made. In this case I believe a
court is likely to find that a making a copy for the purpose of
distribution is an act of distribution and thus the GPL license
applies to the making and subsequent distribution of that copy.

Isaac

Lee Hollaar

unread,
Jul 24, 2004, 5:20:26 PM7/24/04
to
In article Isaac <is...@latveria.castledoom.org> writes:
> I don't see any words in the statute requiring
>transfer of a tangible object.

17 USC 109:
Notwithstanding the provisions of section 106(3), the owner of a
PARTICULAR COPY or phonorecord lawfully made under this title, or
any person authorized by such owner, is entitled, without the
authority of the copyright owner, to sell or otherwise dispose
of the possession of THAT COPY or phonorecord. ...

17 USC 101:
"Copies" are MATERIAL OBJECTS, other than phonorecords, in which


a work is fixed by any method now known or later developed, and
from which the work can be perceived, reproduced, or otherwise
communicated, either directly or with the aid of a machine or device.

> I would expect first sale to apply to


>software that I paid for and received by downloading from the Internet
>as long as I was willing to delete my own copy when transfering it to
>another (ignoring for now the question of whether I owned rather than
>licensed the software).

Maybe, maybe not. Depends on how you do the subsequent transfer. If
you transfer a CD that you have lawfully made from the download, you
fit the language of Section 109. If you send it over the network,
you likely don't.

>There may be a loophole in the literal wording of the statute, but I
>don't believe a court will necessarily interpret the words the way
>Terehkov suggests if it believes that doing so is contrary to the
>intent of Congress.
>
>IMO the statute (17 USC 109) does contain a buglet that could allow
>distribution of fair use copies, computer backups, and other stuff,
>against the will of the copyright holder. I don't believe a court
>would allow the bug to be exploited. I believe instead that the
>court will consider what was done with the copies when deciding
>whether the copies were lawfully made. In this case I believe a
>court is likely to find that a making a copy for the purpose of
>distribution is an act of distribution and thus the GPL license
>applies to the making and subsequent distribution of that copy.

The Copyright Office has noted the same problem. See pages 155-157
in its DMCA Section 104 Report:
http://www.copyright.gov/reports/studies/dmca/dmca_study.html

The report has a good discussion of the problems with first sale
and electronic distribution.

Isaac

unread,
Jul 24, 2004, 6:34:13 PM7/24/04
to
On Sat, 24 Jul 2004 21:20:26 +0000 (UTC), Lee Hollaar
<hol...@faith.cs.utah.edu> wrote:
> In article Isaac <is...@latveria.castledoom.org> writes:
>> I don't see any words in the statute requiring
>>transfer of a tangible object.
>
> 17 USC 109:
> Notwithstanding the provisions of section 106(3), the owner of a
> PARTICULAR COPY or phonorecord lawfully made under this title, or
> any person authorized by such owner, is entitled, without the
> authority of the copyright owner, to sell or otherwise dispose
> of the possession of THAT COPY or phonorecord. ...

Okay. I suppose my language was not clear.

I was questioning whether the owner of the copy is required to have
something tangible given to him by the copyright holder before he
can transfer a copy under first sale. I don't see anything in the
statute requiring that.

As you've posted elsewhere the copy must be lawfully made. It does not
necessarily have to be transfered to the owner on a physical medium. Having
the copy prepared with the authorization of the copyright holder is enough.
It's possible that some unauthorized copies may fit the bill if the
circumstances suggest that they are lawful made.

> 17 USC 101:
> "Copies" are MATERIAL OBJECTS, other than phonorecords, in which
> a work is fixed by any method now known or later developed, and
> from which the work can be perceived, reproduced, or otherwise
> communicated, either directly or with the aid of a machine or device.

Isaac

Lee Hollaar

unread,
Jul 24, 2004, 6:58:26 PM7/24/04
to
In article <slrncg5ov4...@latveria.castledoom.org> Isaac <is...@latveria.castledoom.org> writes:
>I was questioning whether the owner of the copy is required to have
>something tangible given to him by the copyright holder before he
>can transfer a copy under first sale. I don't see anything in the
>statute requiring that.
>
>As you've posted elsewhere the copy must be lawfully made. It does not
>necessarily have to be transfered to the owner on a physical medium. Having
>the copy prepared with the authorization of the copyright holder is enough.
>It's possible that some unauthorized copies may fit the bill if the
>circumstances suggest that they are lawful made.

That's how I understand the statutory first sale provision.

Stefaan A Eeckels

unread,
Jul 24, 2004, 6:49:15 PM7/24/04
to

Indeed, but incorporating (part of) a work in one's
own work is not allowed by first sale. You can sell,
destroy, mutilate the copies you own, but taking
(effectively, making a copy of) a source file from
a copyrighted work and combining it with one's own
code is just not on. Hell, it's what the whole idea
of copyright is all about. It's not because I bought
a magazine that I can start copying parts of articles
in a manual I'm writing. You seem to suggest it's OK
as long as the magazine was downloaded. The simple
fact is that once you start extracting information
from a file, by necessity you have to make copies,
and these are not allowed. The only copies you can
make are those necessary to _run_ a program, and
using a source file in a development isn't part of
running the program you downloaded.

Stefaan A Eeckels

unread,
Jul 24, 2004, 7:00:27 PM7/24/04
to

This is how I understand his argument. I can agree with
Alexander that he can dispose of a file he downloaded as
he pleases, as long as it remains on the medium on which
it was downloaded - first sale pertains to the physical
copy one owns, and doesn't allow the making of further
copies.

As the law clearly considers that loading a program in
RAM is making a copy, and clearly allows all the copies
needed to run a program, it is equally clear that any
other copy is not allowed. Hence, extracting a file from
a tarball, and copying it into a development directory
would be a violation of copyright. A further copy made
by compiling and linking for distribution purposes are
yet other violations.

Using the analogy of a book, one can tear pages out of
a book, but one cannot copy those pages. For a downloaded
tarball, this would mean that one can unpack the tarball
on the same medium one downloaded it on (the paper of a
book), but not make a copy on another medium to distribute
it - one would have to sell the copy and the medium on
which one downloaded it to maintain the analogy.
What Alexander describes is more like copying a CD and
selling the copy.

Barry Margolin

unread,
Jul 24, 2004, 9:55:45 PM7/24/04
to
In article <4102C21A...@web.de>,
Alexander Terekhov <tere...@web.de> wrote:

I'm confused. Your earlier reply implied that you received the CD, not
that you put the copies on the CD after transfering them, because you
contrasted it with the example I gave. If you put it on the CD, then
the CD is "the disk drive you put the copy on" that I referred to.

Barry Margolin

unread,
Jul 24, 2004, 10:03:09 PM7/24/04
to
In article <20040725010027...@DELETEMEecc.lu>,

Stefaan A Eeckels <te...@DELETEMEecc.lu> wrote:

> As the law clearly considers that loading a program in
> RAM is making a copy, and clearly allows all the copies
> needed to run a program, it is equally clear that any
> other copy is not allowed. Hence, extracting a file from
> a tarball, and copying it into a development directory
> would be a violation of copyright. A further copy made
> by compiling and linking for distribution purposes are
> yet other violations.

Where things get ridiculous is when Alexander combines this
interpretation of first-sale with the permissions given in the GPL. It
essentially allows all these copies on your development machine. So his
claim is that since all these copies are lawfully made, he can then
dispose of them all with no restrictions.

Alexander Terekhov

unread,
Jul 26, 2004, 3:41:50 AM7/26/04
to

Barry Margolin wrote:
[...]

> interpretation of first-sale with the permissions given in the GPL. It
> essentially allows all these copies on your development machine. So his
> claim is that since all these copies are lawfully made, he can then
> dispose of them all with no restrictions.

Exactly. That's why the CPL (vs GPL) is a contract, not a bare license.
BTW, I've changed my mind. Even derivative works can be kept "closed
source" (you can "dispose" only the object code [with the GPL attached,
if you like] and withheld the sources) under the GPL and the first sale.

regards,
alexander.

Alexander Terekhov

unread,
Jul 26, 2004, 4:42:39 AM7/26/04
to

Barry Margolin wrote:
>
> In article <4102C21A...@web.de>,
> Alexander Terekhov <tere...@web.de> wrote:
>
> > Barry Margolin wrote:
> > [...]
> > > > > I think that's one of the things most of us disagree with. You don't
> > > > > "own" anything when you make a copy, except maybe for the disk drive you
> > > > > put the copy on.
> > > >
> > > > Or the CD that contains the entire distribution. If you insist, I can
> > > > refrain from media-less distribution, no problem with that.
> > >
> > > But for first-sale to be applicable, you would have to transfer the CD
> >
> > No. I just need to prove that copies that reside on the CD were
> > lawfully made. Making them in the course of authorized downloading
^
|
by me. ----------+

> > is lawful, for example. And I don't think that the law forbids me
> > to store multiple works in an archive/tar (or some "executable")
> > on a CD (or whatever "tangible") and sell it.
>
> I'm confused. Your earlier reply implied that you received the CD, not

No. I've "made" the CD.

> that you put the copies on the CD after transfering them, because you
> contrasted it with the example I gave. If you put it on the CD, then
> the CD is "the disk drive you put the copy on" that I referred to.

Right.

regards,
alexander.

Alexander Terekhov

unread,
Jul 26, 2004, 5:07:52 AM7/26/04
to

Stefaan A Eeckels wrote:
[...]
> Indeed, but incorporating (part of) a work in one's
> own work is not allowed by first sale.

It's not really hard to grasp it. As long as the copy itself is non-
infringing (was not "stolen"), first sale limitation on the exclusive
distribution right shall apply to it. That's the intent of the US
congress and the logic behind various cases in the EU (one was the
decision of BGH against MS in Germany... stating that very cheap OEM
Windows versions can be resold without any hardware by a party that
was not a subject to MS OEM contract restriction... see the light?).

regards,
alexander.

Stefaan A Eeckels

unread,
Jul 26, 2004, 9:58:37 AM7/26/04
to

But in all these cases, the objective is to allow people
to dispose of their physical property, such as the cheap
Windows CDs they lawfully acquired with their PCs, not to
engage in plagiarism.

Let's consider
<http://java.sun.com/docs/books/tutorial/essential/threads/clock.html>.
It says at the bottom:
Copyright 1995-2004 Sun Microsystems, Inc. All rights reserved.

If I click on the link, I get what I believe to be a
legal copy of the article on my screen. There is no
license to accept, exactly like when I download a copy
of a GPLed work, or one of Dan Bernsteins programs.
Can I use the text on that page in my own book on Java?

Stefaan A Eeckels

unread,
Jul 26, 2004, 10:10:18 AM7/26/04
to

Actually, the GPL has nothing to do with it as it only
applies when you want to make copies, prepare derivative
works, and distribute the work. First sale limits the
control the copyright holder has over lawful copies.
The lawful copy is the one that was downloaded, but
doesn't grant you the right to make further copies.
Obviously, once you accept the GPL (which you need to
do to make copies from the downloaded work), you cannot
retro-actively claim that because they are lawful,
the license you had to accept to make them lawful no
longer applies. That's having your cake and eating it.

If you want to invoke first sale, the only thing that
matters is that you posses a lawful copy that did
not require the signature of a contract or the acceptance
of a license to obtain.

Do you have to sign the CPL before you can lawfully
download a copy of the CPLed work?

Alexander Terekhov

unread,
Jul 26, 2004, 12:42:07 PM7/26/04
to

Stefaan A Eeckels wrote:
[...]
> Do you have to sign the CPL before you can lawfully
> download a copy of the CPLed work?

It's covered by things like "catch all" terms and conditions of using
the Eclipse web site, ftp servers, mail lists, etc. OSI's counsel
wrote a nice paper on "manifestation of assent", BTW.

regards,
alexander.

Alexander Terekhov

unread,
Jul 26, 2004, 12:51:23 PM7/26/04
to

Stefaan A Eeckels wrote:
[...]
> But in all these cases, the objective is to allow people
> to dispose of their physical property, such as the cheap

And I sold my CD -- physical property. Cheaply. ;-)

> Windows CDs they lawfully acquired with their PCs, not to
> engage in plagiarism.

You don't get it. I can create a million+1 copies and/or derivative
works (all kept as trade secrets without any sources, on many CDs)
of the GPL'ed stuff and no one could claim infringement. Now, if I
want to *distribute* those copies (as "physical property"), the
first sale says that I can do it without any authorization of the
copyright owner (s/he can't "condition" that act without some
binding contract).

The Munich district court ruled this month that the GPL should be
treated as a contract (hello Prof. Moglen) but, given the context
(confirmation of a totally harmless preliminary injection without
any arguments against the GPL raised by the Sitecom), they simply
briefly looked at some isolated sections of the GPL and failed to
realized that the GPL itself says that it is based purely on the
copyright law (section 5 bullshit***). The fact is that the GPL
is preempted by the copyright law, so to say. Imagine that you
buy a CD full of the GPL'ed binaries, the seller mails you that
funny "written offer", and your dog eats it the next day... do
you believe that now you can't resell your CD without that piece
of paper (without risking to go to jail)?

>
> Let's consider
> <http://java.sun.com/docs/books/tutorial/essential/threads/clock.html>.
> It says at the bottom:
> Copyright 1995-2004 Sun Microsystems, Inc. All rights reserved.
>
> If I click on the link, I get what I believe to be a
> legal copy of the article on my screen.

You don't own it, just like you don't own pictures on your TV
screen. SUN's web site is "performing", IIUC. No copy gets
distributed or authorized to be made and owned by you.

regards,
alexander.

***) "nothing else grants you permission to modify or distribute
the Program or its derivative works." It appears that the FSF
never heard of Section 109 and Section 117.

Arnoud Galactus Engelfriet

unread,
Jul 26, 2004, 3:04:34 PM7/26/04
to
In article <4105368B...@web.de>,

Alexander Terekhov <tere...@web.de> wrote:
>The Munich district court ruled this month that the GPL should be
>treated as a contract (hello Prof. Moglen) but, given the context
>(confirmation of a totally harmless preliminary injection without
>any arguments against the GPL raised by the Sitecom), they simply
>briefly looked at some isolated sections of the GPL and failed to
>realized that the GPL itself says that it is based purely on the
>copyright law (section 5 bullshit***).

Given that German law considers giving someone directions to
create a contract, it's not surprising that they would treat
the GPL as a contract.

Arnoud

--
Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself
Patents, copyright and IPR explained for techies: http://www.iusmentis.com/

Alexander Terekhov

unread,
Jul 26, 2004, 3:33:30 PM7/26/04
to

Arnoud Galactus Engelfriet wrote:
>
> In article <4105368B...@web.de>,
> Alexander Terekhov <tere...@web.de> wrote:
> >The Munich district court ruled this month that the GPL should be
> >treated as a contract (hello Prof. Moglen) but, given the context
> >(confirmation of a totally harmless preliminary injection without
> >any arguments against the GPL raised by the Sitecom), they simply
> >briefly looked at some isolated sections of the GPL and failed to
> >realized that the GPL itself says that it is based purely on the
> >copyright law (section 5 bullshit***).
>
> Given that German law considers giving someone directions to
> create a contract, it's not surprising that they would treat
> the GPL as a contract.

That ruling is quite funny, isn't it? They've simply took the
FSF's "unofficial translation" with hits like "darauf basierendes
Datenwerk" (uhmm, derivative work... Bearbeitung und Umgestaltung,
I mean), didn't bother to include the German translation of the
GPL Section 2 at all (they took only sections 1, 3, and 4) and
yet wrote that both sections *2* and 3 seem to be OK, so to say.
Given other flops and typos... too much Bier, I guess. ;-) ;-)

regards,
alexander.

Alexander Terekhov

unread,
Jul 26, 2004, 6:05:49 PM7/26/04
to

Isaac wrote:
[...]
> While I don't subscribe to Terekhov's first sale theory, ...

I like this:

http://google.com/groups?selm=slrna7gh48.6s.isaac%40latveria.castledoom.org

<quote authors=Jeffrey Siegal, Isaac>

>What about the first sale doctrine? Indeed, if users own their own
>copies, including binary copies, of the software, than those users can
>transfer those copies without complying with the GPL's requirements as
>to source code access.

Interesting. I really hadn't thought about that and my attempts
to digest the implications on the fly are making my head hurt.
I can't resolve the problems introduced by first sale without
either making the user of GPL'd code a non owner or deciding
that users really can transfer their copies, no matter what the
GPL says.

Gotta think about this. Kinda makes my whole argument moot
if GPL users are really licensees. <g>

Isaac

</quote>

Uhmm, <q> ;-)

regards,
alexander.

Isaac

unread,
Jul 26, 2004, 7:30:42 PM7/26/04
to

I believe my theory as expressed in that previous post to be
incorrect based on some discussion on first sale involving fair use
copies in anohter forum.

The loophole in the statues does exist. In theory, the statute
would allow distribution of backup copies of computer software as
made under 17 USC 117 as well as fair use time shifted copies of
TV broadacasts, but I believe a court would easily find that Congress
did not intend to allow those things under first sale.

Isaac

Stefaan A Eeckels

unread,
Jul 26, 2004, 5:42:55 PM7/26/04
to
On Mon, 26 Jul 2004 18:51:23 +0200
Alexander Terekhov <tere...@web.de> wrote:

>
> Stefaan A Eeckels wrote:
> [...]
> > But in all these cases, the objective is to allow people
> > to dispose of their physical property, such as the cheap
>
> And I sold my CD -- physical property. Cheaply. ;-)

The fact that you own the paper doesn't mean you
can photocopy a book onto it, then sell the paper.

> > Windows CDs they lawfully acquired with their PCs, not to
> > engage in plagiarism.
>
> You don't get it. I can create a million+1 copies and/or derivative
> works (all kept as trade secrets without any sources, on many CDs)
> of the GPL'ed stuff and no one could claim infringement.

As long as it's for your own use, or no-one knows, indeed.

> Now, if I
> want to *distribute* those copies (as "physical property"), the
> first sale says that I can do it without any authorization of the
> copyright owner (s/he can't "condition" that act without some
> binding contract).

You can sell the copies of the original work. This is
what the lads from CheapBytes do, for example. And as
the original work contains the source, it happens to
comply with the GPL as well.

Why you think that after preparing a derivative work
(which you specifically cannot do, as per the copyright
statutes, unless it is done to run a program) you could
still claim first sale, escapes me. You simply have no copy
of the derivative work that you lawfully obtained.

> Imagine that you
> buy a CD full of the GPL'ed binaries, the seller mails you that
> funny "written offer", and your dog eats it the next day... do
> you believe that now you can't resell your CD without that piece
> of paper (without risking to go to jail)?

In the case of an isolated incident, probably not. In any
case, it suffices to make the buyer aware of the offer
by your original seller.

> > If I click on the link, I get what I believe to be a
> > legal copy of the article on my screen.
>
> You don't own it, just like you don't own pictures on your TV
> screen. SUN's web site is "performing", IIUC. No copy gets
> distributed or authorized to be made and owned by you.

Funny, it's sitting right there on my disk, because it
happens to appear even when the machine is off-line.

Is there a difference between a file that is downloaded
for display with a browser, and a file that is downloaded
(like qmail-1.03.tar.gz) and looked at with gtar?
The instructions given to the web server are identical,
as wget so aply illustrates. When do you "own" a copy
of a file you downloaded?

BTW, section 117 says:

| (b) Lease, Sale, or Other Transfer of Additional Copy or Adaptation. -
|
| Any exact copies prepared in accordance with the provisions
| of this section may be leased, sold, or otherwise transferred,
| along with the copy from which such copies were prepared,
| only as part of the lease, sale, or other transfer of all
| rights in the program. Adaptations so prepared may be transferred
| only with the authorization of the copyright owner.

In other words, you cannot distribute the derivatives,
even though you can prepare them for your own use.
Other lawful copies can only be transferred with the
original copy. There is no first sale here, as there
was no sale, just you making a lawful copy. Lawful copies
you obtained are not lawful copies you made, legal eagle.

You clearly got your knickers in a knot. In addition,
your selective answering and constant evading of the
issues make further discussions pointless. Enjoy your
fictional world.

Alexander Terekhov

unread,
Jul 27, 2004, 3:49:51 AM7/27/04
to

Isaac wrote:
[...]

> The loophole in the statues does exist. In theory, the statute
> would allow distribution of backup copies of computer software as
> made under 17 USC 117 as well as fair use time shifted copies of
> TV broadacasts, but I believe a court would easily find that Congress
> did not intend to allow those things under first sale.

Backups aside for a moment, fair use doesn't create "lawfully made
copies", IIUC. The copies are infringing but the "use" is fair and
doesn't create liability (as long as it's fair).

http://www.chillingeffects.org/dmca512/notice.cgi?NoticeID=320
http://google.com/groups?selm=c29b5e33.0204090425.37611f83%40posting.google.com

;-)

regards,
alexander.

Alexander Terekhov

unread,
Jul 27, 2004, 4:03:23 AM7/27/04