🚀 go-pugleaf

Since there does seem to be some interest, I've collated a few items relating
to my agrument that GPL is unenforceable.  This first item is an article I
posted back in January:

==============================================================================
Newsgroups: alt.religion.computers,gnu.misc.discuss
Subject: Re: Disinfecting the GNU Public Virus...er...License
Summary: GPL 2(b) is unenforceable
References: <4&VSZ:@splut.conmicro.com>
	    <NELSON.90Jan16164120@image.clarkson.edu>
	    <SJA.90Jan17142417@sirius.hut.fi>
	    <NELSON.90Jan18165545@image.clarkson.edu>
	    <SJA.90Jan20170637@sirius.hut.fi>
Sender:
Reply-To: kozma@rex.UUCP (John Kozma)
Followup-To:
Distribution:
Organization: Computer Science Dept., Tulane Univ., New Orleans, LA


I guess I'm still a bit sheepish since someone pointed out to me, in response
to some of my postings last summer on copyright law, that copyright notice is
no longer mandatory on published works, as the US is now a signatory to the
Berne treaty.  I don't think this fact undermines my agrument, however, that
section 2(b) of the GNU Public License is unenforceable, at least to the
extent that it purports to require the author of a program to relinquish
any rights in such program.

As I understand the intent of the license, it allows me to distribute copies
of, say, GNU library routines (G) linked with my object code (MO), so long
as I agree to supply the distributee with my source (MS).  The problem is
this: I had a copyright in MS before I ever linked it with G, and as the
owner of that copyright, I have the exclusive right to make and distribute
copies of MS.  By subjecting MS to the GPL, I would effectively be giving
up these rights.  (Yes, I can still make and distribute copies of MS myself,
but the exclusive right which constitutes ownership of a copyright means
the right to exclude others from making or distributing copies, which I
no longer have, because I must allow my distributee to copy and further
distribute, and he in turn must allow the same for his distributees, etc.)
I think it has been correctly pointed out that being subject to GPL is not
equivilant to being in the public domain.  Thus, somebody must have exclusive
rights for MS, even if I don't, and the only possibility is that the owner
of the copyright to G, my licensor, got them.

Thus, GPL 2(b) purports to transfer ownership of my copyright in MS to the
owner of G, upon my distributing a copy of MO linked to G.  But this is not
possible under section 204(a) of the copyright law (Title 17 of the US Code),
which states: "A transfer of copyright ownership...is not valid unless an
instrument of conveyance, or a note or memorandum of the transfer, is in
writing and signed by the owner of the rights conveyed...."

I may try to keep up with this group to see if there are any followups to
this post, but I try to be at least as lax about the usenet as I am about
my dissertation, so if you have questions about my argument, I'd suggest
e-mail.  (If there are enough questions/criticisms, I will post a clarified/
modified argument.  Please no suggestions that I consult with an intellectual
property lawyer, though.  I am one.)

John P. Kozma
kozma@rex.cs.tulane.edu
==============================================================================

Following this post, I exchanged several messages via e-mail with a Mike
Haertel, wherein I proposed that the right to redistribute which must be
granted under the GPL is actually revocable because it is not supported by
consideration, and he disagreed.  His argument was that there is a duty not
to revoke the license, and that this duty is owed to the FSF under the GPL.
I did not respond further, but to me his argument is obviously flawed because
it ignores my original argument.  This is the last message Mike sent me:

==============================================================================
Date: Wed, 31 Jan 90 14:32:36 -0600
From: mike@umn-cs.cs.umn.edu
Subject: Re: Further question on GPL
To: KOZMA@comus.cs.tulane.edu
Message-Id: <9001312032.AA27600@umn-cs.cs.umn.edu>

>     If you wish to remain in compliance with the GPL, then any code you
>     distribute linked with GNU code must be under terms that allow the
>                                                 ^^^^^^^^^^^^^^^^^^^^
>     distributee to further redistribute.
>     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Thus, I must grant my distributee a license to further distribute my code.
> This is okay so far, but I would argue that this license is freely
> revocable by me at any time, since it is not supported by consideration

No.  See section 6.

> (which the GPL makes quite clear).  You continue:
>
>                                         If you later decided that you
>     wished to make a separate distribution of your code, you could place
>                    ^^^^^^^^^^^^^^^^^^^^^
>     any restrictions on copying that you chose, provided that you did not
>     include any GNU software in your restricted distribution.
>
> My right to DISTRIBUTE my code myself is not really in question.  My copyright
> allows me to EXCLUDE OTHERS from distributing.

Yes.  However, their right to redistribute, once granted as required by
the GPL, may not be revoked or you will be in violation of the GPL.
The notion of a "consideration" has nothing to do with it.

>                                                              However,
>     people who had received your original free distribution would still be
>                                                                 ^^^^^^^^
>     able to copy from that.
>     ^^^^^^^^^^^^^^^^^^^^^^
>
> Under their license from me, yes, (or insofar as is necessary under the
> copyright law to use my code, which probably doesn't allow them to copy my
> source code at all), but they can't DISTRIBUTE further EXCEPT under my
> license, which as I've explained above may be revoked by me at any time.

Which I've explained may *not* be revoked, once granted, or you will
be in violation of the GPL.

>                            But, in distributing your code linked with
>     GNU code you give up neither your copyright nor your future right
>     distribute your code again under different terms.
>
> By the analysis above, I agree.  But if the license I grant to my distributee
> is not freely revocable by me at any time by my whim, then I HAVE given up
> my right to exclude others from distributing copies of my source code, and,
> as such, a significant part of my copyright.
	      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

This is a matter of interpretation.  You have given up your right to control
copying and redistribution of your code as obtained from the specific
distribution you made with GNU code.  However, if you distribute your work
again without GNU code, you can restrict it any way you please.  So, yes
you have given up "a significant part of your copyright", with respect to
copies of your work derived from one particular source.  In general, you
have not given up any part of your copyright.

> Such a license would pass
> muster under 17 USC 204(a), in my opinion, but would not be effective under
> the law of contracts (except perhaps in Louisanna) because, as I've pointed
> out, there is no consideration.

Sorry, you're considering only half the picture.  Attempting to revoke
your receipients rights would place you in violation of your agreement
with us.

> Of course, I can only revoke my license as it applies to my code.  I have
> no authority to revoke the GPL as it applies to works the copyrights to which
> are owned by someone else (eg, GNU).
>
> I hope we can continue this dialogue, and I appreciate your invitation to
> post your explanation, as I think this is one of the major sources of
> disagreement and misunderstanding in the GPL.  Is my theory of a revocable
> license reasonable?

No.
==============================================================================

Though I do not expect to persuade any true believers in the GPL, I offer
some further comments in support of my argument.

To frame the question precisely, I think
one must compare the standard kinds of license provisions-- exclusive and
non-exclusive, of which the GPL is neither.

An exclusive license, which
is a transfer of copyright ownership if unlimited as to time and territory
(which the GPL is), means that the licensor gives the licensee not only the
right to distribute (or make copies, or exercise one of the other exclusive
rights of copyright), but further that the licensor promises that he will
not grant the right to distribute (etc) to anyone else.  That is, the right
to distribute (etc) is granted exclusively to the licensee.  Once this is
done, (even if the licensor has reserved to himself the right to distribute
but has limited his right to let others distribute), it is the licensee
who has the exclusive right, ie, the right to exclude others from distributing
(or to grant that right in return for a royalty).  The one who has the
exclusive right is, in effect, the copyright owner, and under 17 USC 204(a)
one cannot become an owner (other than as an author, of course, or in the
case of works made for hire) unless there is a signed instrument by the
previous owner transferring ownership.

A non-exclusive license is a license
to distribute (etc) without an accompanying promise by the licensor not to
grant further licenses: in this case it is the licensor who retains the
exclusive right, ie the right to grant or deny further licenses.  It is
possible to grant a public license, but if such a license is not a dedication
(placing the work in the public domain), it must be freely revocable by the
copyright owner (for any reason, or for no reason).  The GPL purports to
require its licensees to, under certain conditions, to grant a license
(which I maintain is a public license, even though Mr. Haertel emphasises
that it only relates to specific distributions) in the GPL licensee's
copyrighted (source) code.  This is fine, but I maintain that such a license
is freely revocable by the grantor, even with respect to his distributee,
because it is unsupported (as required by the GPL) by consideration.  If
the GPL licensee somehow owes a duty to FSF (the GPL licensor) not to revoke
the public license he granted his distributee in his (the GPL licensee's)
code, then he (the GPL licensee) effectively shares the right to exclude
others from distributing (etc) his (the GPL licensee's) code, which means
the FSF has obtained part ownership of his (the GPL licensee's) copyright.

Well, if you followed all that, you see that I don't think my original
argument has been answered.  I would note, though, that the GPL may be
strengthened by consideration of the authors moral rights which have
presumably been expanded by the US adoption of the Berne treaty.