[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: PHP-Nuke License Conclusion?

On Tue, May 13, 2003 at 02:35:05PM +0200, Henning Makholm wrote:
> Scripsit Branden Robinson <branden@debian.org>
> > On Tue, May 13, 2003 at 06:46:31AM +0200, Henning Makholm wrote:
> > > Digging in the archives turns up that it has not always been you who
> > > made the false claim that GPL+more restrictions is necessarily
> > > internally inconsistent. I apologize for implying that.
> > Calling it a "false claim" is overreaching given that you've never
> > actually constructed an argument illustrating its falsity.
> I wonder how the arguments I pointed to came into being, then, if I
> did not construct them.

Which arguments?  You keep saying they exist, and keep failing to cite
them, whether by reference or inclusion.

> > Okay, so you and Nick Phillips will cotton to "the GNU GPL" being a
> > variable rather than a constant,
> Strawman.

How so?  You and Nick are claiming that the "terms and conditions" of
the GNU GPL as applied to a work are not always the literal text of what

Thus, the "terms and conditions" -- even in written form -- of the GNU
GPL as applied to a given work can vary, even when the written text
MODIFICATION" and "END OF TERMS AND CONDITIONS" does not in fact change.

> Your only point seems to be that *sometimes* the description of
> such almost-but-not-quite-GPL licensing terms is phrased in unclear
> and possibly inconsistent ways.

No, that's an incidental problem.  My point is that whatever a license
is that is the GNU GPL plus additional restrictions (consistent or not),
it isn't the GNU GPL.

> This in no way entails that *every*
> set of almost-but-not-quite-GPL licensing terms must be unclear and
> inconsistent.

Where the set includes restrictions on activity not enumerated between
"END OF TERMS AND CONDITIONS", that's correct.  This is because the text
of the GNU GPL is self-referential, and talks about "these terms and
conditions" (clause 6).

> >   [I have asserted that the way most people add restrictions to the GNU
> >   GPL] -- at least in cases that have come to my attention via the
> >   debian-legal list -- results in a license that is internally
> >   inconsistent, and thus impossible to satisfy while exercising one's
> >   traditional freedoms with a copyleft.
> Even here, you seem to argue that because "the way most people add
> restrictions" is wrong, and often internally inconsistent - which I do
> not deny, mind you - it is impossible to add restrictions.

If they were to modify the GPL text, they could achieve what they desire
in a (potentially) self-consistent way.  Most people don't bother to do

> >   Observation One
> >   ---------------
> >   The scope of the terms and conditions, a.k.a. restrictions, is clearly
> >   defined by the text of the GNU GPL.  Any restrictions beyond that are
> >   forbidden by section six.
> Section six has absolutely no power over how an individual, sole,
> author may or may not chose to license his work.

What the author chooses and what he manages to express are not always in
consistent harmony.  Copyright licensees are bound first and foremost by
the licensor's explicit terms and conditions.  Licensees who operate
under a "gentlemen's agreement" with the licensor in which they violate
the literal terms of the license do so at their own peril.

I do not think the Debian Project should encourage its users to assume
that they are the beneficiaries of a "gentlemen's agreement" with the
licensor, when no such agreement was actually formed.

> >   It is obvious to me that "THESE TERMS AND CONDITIONS" is meant to refer
> No.

Well, then, I suspect we have an unbridgeable gulf here.

> On the contrary, it is obvious that if an author knowingly and
> explicitly licenses is work under "the conditions of the GNU GPL, with
> the following additional restrictions: [bla bla]", then the author
> intends every occurence of "THESE TERMS AND CONDITIONS" to refer to
> his entire combined license conditions.

You're drawing an inference.  It is not obvious.  When an author uses
the GNU GPL, he is speaking with his own voice, not the Free Software

When an author says to me:

Copyright 2003 Joe Blow.


  1) You can do foo.
  2) You can do bar.
  3) You can do baz.
  4) Each time you redistribute the Program (or any work based on the
  Program), the recipient automatically receives a license from the
  original licensor to copy, distribute or modify the Program subject to
  these terms and conditions. You may not impose any further restrictions
  on the recipients' exercise of the rights granted herein. You are not
  responsible for enforcing compliance by third parties to this License.


  You can't do baz.

...then I become confused because he is simultaneously telling me I can
and cannot do "baz".

> >   Observation Two
> >   ---------------
> >   The FSF claims copyright in the text of the GPL itself, and does not
> >   extend permission to anyone to modify it.
> Actually, the FSF does permit people to use the legal language in the
> GPL as the baseline of other licenses (see your own quote from the GPL
> FAQ below). They will not like it, and they will call people who do it
> nasty names, but they do not claim the legal power to prevent it.

Actually, they do claim the legal power to prevent it.

"Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed."

I agree that there is some evidence that the FSF has softened on this
point over the past 12 years.  So, if people really can take advantage
of the open season on the Terms and Conditions of the GNU GPL, who don't
we make sure that licensors stop talking out of both sides of their

> >   Therefore, making modifications to the area between "TERMS AND
> >   AND CONDITIONS" might be considered by the FSF as an infringement of
> >   their copyright in the text of the GNU GPL.
> Wrong (see above, and below).

I have not seen a statement from the FSF repudiating its assertion of
copyright in the entire text of the GNU GPL, nor granting blanket
permission to the public to modify the terms and conditions of the GNU
GPL to its taste.

> But even if *making* the modifications were illegal, it is beyond the
> powers of the FSF to forbid people to *imagine* modifications, and use
> such imagination as a succint way to describe their own licensing
> terms.

People imagine logical contradictions and paradoxes all the time.  Ever
been to Sunday School?

That such things can be imagined does not mean that they are useful for
copyright licenses, which should be clear, unambiguous, and lawful.

> >   Observation Three
> >   -----------------
> >   The FSF does not want people "forking" the GPL, and think that license
> >   proliferation is a bad thing.
> This is true, but it does nothing to argue that a fork would be
> internally inconsistent.

And I have not made such an argument.  People can fork the terms and
conditions of GNU GPL all they want as far as I'm concerned, but if they
create a license document that is self-contradictory I must insist that
the Debian Project obtain clarification, and if that fails we must
reject it.  Our users must be able to trust us to apply logical
reasoning in interpretation of license documents.  If we countenance
paradoxical licenses as DFSG-free, we undermine that trust.

> > 	Although we will not raise legal objections to your making a
> > 	modified license in this way, we hope you will think twice and not
> > 	do it.
> Note this excerpt from your own FSF quote.

Okay.  I've noted it.  Why am I noting it?

> >   However, in practice, most people don't do what the FSF recommends in
> >   their FAQ.  They just say "it's under the GPL" and slap extra
> >   restrictions somewhere else -- perhaps in a README file.
> Which I can easily agree is internally inconsistent and may make the
> work undistributable.

Okay.  You can easily agree -- but you're not going to???

Color me confused.

> But this still does not mean that adding additional restrictions is OF
> NECESSITY internally inconsistent.

It is if they use the GNU GPL's terms and conditions as-is.

> >   It is this practice that I call internally inconsistent,
> The words you used were
> || Debian interprets "this License" and "herein" to mean the
> || conditions of the GNU GPL expressed in its text; no more and no
> || less.
> which does not seem to me to acknowledge that there is a right way to
> add extra restrictions. (Right as in "probably legally sound", not
> right as in "free").

Of course there is.

> >   And, of course, even if one feels that this isn't a
> >   problem, there is *still* the issue of incompatibility with all other
> >   GPLed works in the world.
> I do not deny this. But an incompatible license can still be
> *internally* consistent.

It will be if they use the GNU GPL's terms and conditions as-is.

> >   I therefore think it's a good idea for people forking the GPL to
> >   rename the license,
> But I still do not think you have offered any convincing arguments
> that "GPL with the following additional restrictions: [blah blah]"
> is a bad name to rename it to.

Well I think I have.  Such a license, whatever it is, is not the GPL,
and will not be GPL-compatible[1].

> On the contrary, I think it is the honest way to go about it, rather
> than reproducing the entire load of legalese as "the Foobar Public
> License"

A license name one hasn't seen before is a damn good reason to read it

> and hiding the incriminating

Eh?  Why are you suddenly alleging criminal activity?

> restrictions in the midst of vaguely-familiar licensing text that
> nobody is going to read closely anyway because it just looks like a
> revamped GPL anyway.

We here on debian-legal will read it closely.  We'd better!  And so
should anyone else who doesn't want to be a copyright infringement

> > To do that, all you have to do is show that the phrase "THESE TERMS
> > AND CONDITIONS" in the GNU GPL is not meant to, and does not in
> > fact, refer to the text between "TERMS AND CONDITIONS FOR COPYING,
> It is clear as day that if the author starts his source files with
> /* Copyright YYYY N.N.
>  * This file can be copyed freely under the GNU general public license
>  * with the added restriction that everyone who distributes it (or
>  * derived works) must pet a cat.
>  */
> then OBVIOUSLY he means for "THESE TERMS AND CONDITIONS" to include
> his additional restriction.

Except elsewhere, when speaking with his own voice, he says between the
"END OF TERMS AND CONDITIONS" that "Activities other than copying,
distribution and modification are not covered by this License; they are
outside its scope."

So the licensor is telling me that he has is not restricting my
non-copying, non-distributing, and non-modifiying activities, and in the
same breath is telling me that other activities *are* within the scope
of his license, and petting cats is one of those activities.

> And it is what the copyright holder means that counts.

Certainly, where that meaning can rationally be deduced from the words
he uses.  But if he hands me a logical paradox and calls it a license,
I'm going to hand it back to him, and so too should the Debian Project,
in my opinion.

> It does not matter what those who happened to write the words that the
> author chose to refer to to partially explain what he intends meant,
> as long as the author's intentions demonstrably differ.

I don't question that they likely differ.  Discerning exactly what he
*does* mean is another matter entirely.

[1] except possibly in pathological cases, which I haven't yet been able
to imagine

G. Branden Robinson                |    I must despise the world which does
Debian GNU/Linux                   |    not know that music is a higher
branden@debian.org                 |    revelation than all wisdom and
http://people.debian.org/~branden/ |    philosophy. -- Ludwig van Beethoven

Attachment: pgpsyjnnCkPXI.pgp
Description: PGP signature

Reply to: