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

Re: KDE not in Debian?

Lynn Winebarger wrote:

> On Wed, 2 Feb 2000, Andreas Pour wrote:
> > Lynn Winebarger wrote:
> >
> > >    Scanning through your posts, all indications are that you refuse to
> > > listen.  It is certainly possible to distribute XFree86 (and any
> >
> > > derivatives) under the GPL or practically any license (as long as it
> > > preserves the copyright notice) under the sublicensing permission.
> >
> > The XFree license also says you have to include the XFree license in any copies you
> > redistribute.  I suppose you might argue there is no reason for that and we can
> > safely ignore that rather critical condition.  In that case it is you that appears
> > not to listen.
> >
>    No.  I'm saying the X license allows other interests to be attached in
> the form of licenses, and the X copyright notice only notifies you of
> their (the copyright holder's) permissions.

Agreed  -- the XFree license applies only to the XFree code.

> > Wrong, it says you need to include the notice with any "substantial portion" of the
> > code.  Notice that this license applies to each *file*, not the entire XFree build
> > tree.  So including any substantial portion of any file (however much that is),
> > requires you to include their license.  And since Debian redistributes basically all
> > of their files untouched, there is no question that Debian has to include the XFree
> > copyright notice and the XFree license (which is referred to as the "permission
> > notice").
> >
>     The notice has to go in.  However, the original authors cannot give
> permissions for modifications of the code - they don't have the sole
> copyright interest in derivative works.  The copyright notice also does
> not claim it is the sole license for the code.

It is the sole license for the XFree code, but not for a derivative work.

>    For example, author A distributes his work W to person B.  Now B gives
> a copy to C with a license that has additional restrictions X, Y, and Z.
> Now, the X license gives the permission of A for person C to exercise the
> rights of the copyright monopoly.  That, however, does not invalidate the
> additional restrictions B has placed on the license for the copy.

I don't see how they are enforceable.  The copyright holder, A, has said C can do certain
things, B can't change what A has permitted C to do.  But in the event this is not clear
enough, XFree code specifically says you can sublicense XFree code, but only if you
include the XFree license.  If that were to mean that B can change the license however B
wants, what would be the point of forcing B to include the XFree license?  This important
ir not critical "condition" -- in fact about the only condition -- to sublicensing is not
going to be read to mean nothing by any fair-minded reader, but your reading precisely
makes it mean nothing.

> Only B
> can waive those restrictions.  I don't see anywhere in the X license
> claiming that it contains the sole terms of the license to C.Quite the

> opposite in fact ("... including without restriction the rights to ...
> sublicense ...").

Right, but the "right" to "sublicense" is subject to the obligation to include the X
Copyright and the X license in the copy B distributes.  Again, if the License means
nothing, and in fact C cannot "deal in the Software without limitation" (as the X license
provides), what is the point in including the license in the copy that's distributed?
Please explain that to me.

>     While I don't know whether a court would enforce such restrictions on
> the unmodified code without the intervention of the author, the X license
> certainly allows such additional restrictions.

I don't see where it does.  The fact that X license does not explicitly prohibit changing
the license does not imply a right to change it.  The fact that the X license
affirmatively requires copies of the code to include the license is more than enough.

>    Here's an example of how such a sublicense might appear:
> --------------------
>    You may not copy, modify, or distribute this software.
>                                          XYZ Software Co.
> Copyright (C) 1996 X Consortium
> Permission is hereby granted, free of charge, to any person obtaining a
> copy of this software and associated documentation files (the "Software"),
> to deal in the Software without restriction, including without limitation
> the rights to use, copy,
> modify, merge, publish, distribute, sublicense, and/or sell copies of the
> Software, and to permit persons to whom the Software is furnished to do
> so, subject to the following conditions:
> The above copyright notice and this permission notice shall be included in
> all copies or substantial portions of the Software.
> Except as contained in this notice, the name of the X Consortium shall not
> be used in advertising or otherwise to promote the sale, use or other
> dealings in this Software without prior written authorization from the X
> Consortium.
> X Window System is a trademark of X Consortium, Inc.
> -------------
>    There is no contradiction.  X grants you their permission, XYZ denies
> you theirs, as is their right under the X license.

I don't agree.  Where does the X license specifically grant the right to alter the
license?  It doesn't --you "imply" this rather powerful right to sublicense under any
terms you want from the mere absence of an explicit prohibition against doing that, all
the time ignoring the purpose of the requirement that the sublicensor include the original

license.  Note, incidentally, that the X license does not require the sublicensor to
include the dislcaimer of warranty, only the License.  Why is this so?

The point can be stated another way.  B has no right to prevent me from copying the X
code.  B is not the copyright holder, A is.  A tells B, you can copy this code and
distribute it, but only if you put my license on it.  B does so, but adds some
restrictions.  C gets the code from B.  C sees where A, the copyright holder, the only
person that can prohibit copying, has said, anybody, including C, can copy the code.  So
why would C care about B's restriction?

Do you really think that if C copied the code, despite B's notice that C can't, then B can

sue C for copyright violation?  Do you really think that?  B can't sue for copyright
violation b/c B has no copyright, right?  That only leaves B to sue for a license
violation, but a copyright license presupposes a copyright, so that won't do B any good
either, as B has no copyright.  So on what basis could B enforce this "added
restriction"?  (Of course, if B can't enforce it, it is not a restriction, it's wishful

> This sublicense won't
> apply to other copies you get (unless they carry the same sublicense, of
> course).
> > > It doesn't invalidate any
> > > other agreements you may entered regarding the code.  It also doesn't
> > > invalidate different licenses on derivative works, particularly those with
> > > sufficient copyrightable content to be protectable.
> >
> > No it does not, and I have never claimed it does.You can license additions to XFree
> > code under your own license, and when someone distributes the combined work they have
> > to comply with the XFree license and your license.  The XFree code, however, at all
> > times remains under the XFree license.
>    No, the permissions explicitly allow you to attach a license to the
> unmodified code.

Where?????  It does explicitly allow you to "sublicense", but it is a far cry to go from
the right to sublicense to go to the right to sublicense under any terms you want.  But,
it's even easier w/ the XFree code, b/c it says you can sublicense only if you include the

XFree license (but not the warranty disclaimer) in the resulting code.

As I have mentioned before, the reason the right to sublicense is explicitly granted is
b/c of the contract doctrine of "privity".  In order for a contract/license to be
effective, there has to be a direct connection between the licensor and the licensee (this

is the point of Section 6 of the GPL, to avoid the sublicensing issue by giving a direct
license from A to C).  By giving B the right to sublicense to C, that means C can enjoy
all the rights of the license w/out having received a license from A.  If B did not have a

right to sublicense to C, then someone could argue (successfully in many jurisdictions)
that C did not have the right to make modifications, etc., since there was no "privity of
contract" between C and A, the copyright holder, and hence the license does not apply to

> You can enter agreements with respect to that code,
> completely unmodified.  You can license the whole of any derivative work
> you create as you want (subject to the minor conditions of the X license).

You cannot really license the derivative work how you want -- but you can copyright your
changes and license those however you want.  Then, to distribute the combined work,
someone must comply with both licenses.  However, the XFree code remains under its own

> The X license allows you to add as many restrictions as you like to a
> work, modified or not,


> but I stop short of saying that since it's hard to
> tell if a court would enforce such a sublicensing if the plaintiff has no
> direct copyright interest in the work.

Hard to say?  Don't think its hard at all, a court very likely would not.  Remember, the
only reason a license is needed is b/c under law no copies can be made.  The copyright
holder has an "exclusive" right to make copies.  With a license, the copyright holder can
allow others to make copies.  Thus, the point of a license is for the copyright holder to
permit copying of what otherwise by law is not permitted to be copied.  Now you come along

and you say B can create a new license which *prevents* copying of what the copyright
holder has said can be copied *by anyone* *without limitation*.  Lynn, you have managed to

turn copyright law on its head in a few clicks of the keyboard.  I, for one, don't buy it.

> For derivative works or cases
> where you enter into a contract, the added restrictions are going to have
> more weight.
> > >  Furthermore, it is
> > > perfectly legal to distribute the exact same expression under multiple
> > > copyright licenses.
> >
> > For the author, sure.  For a recipient under a particular license, absolutely not.
> > Otherwise I could take my MS Word and distribute it under the GPL, couldn't I?
> >
>     No, but MS Word isn't accompanied with a copyright license that allows
> you to do so.  XFree86 _is_ accompanied by such a license.


> > > I don't have to "re-license" all currently existing
> > > versions of the code to comply with the GPL on a version that includes or
> > > is derivative of some GPL'ed work.  I only have to distribute that version
> > > under the GPL.
> >
> > What does that mean?  You can't distribute any version of the XFree code (i.e., if
> > you copy a "substantial portion" of it) under the GPL, it remains under the XFree
> > code license.
> >
>    Well, that's what's under dispute, isn't it?
>    It means the license that
> accompanies one copy of a work has no effect on a license that accompanies
> a second copy of the very same work.
> > > If you insist on continuing to debate what many of us consider obvious
> > > (e.g. that X licensed code may be redistributed under a proprietary
> > > license) then it would be best for you to post (in a relatively short way)
> > > your basic assumptions that are leading you to your conclusions.  My wager
> > > is that we'll disagree with one of your assumptions, so maybe we'll get
> > > somewhere more fruitful (no guarantees though).
> >
> > Well:
> >     (a) copyright law prevents copying of protected works without permission from the
> > copyright holder;
>    True.
> >     (b) that permission to copy can be given in a document, whether it is called a
> > "license" or a "permission notice" or whatever, so long as in substance it permits
> > copying;
>    True.
> >     (c) if someone grants such permission, you can only copy in compliance with the
> > grant of permission;
>    True but incomplete - there may be other conditions attached besides
> the original granter's permission.

Not if the work remains unchanged, see above.

> [deletia - mostly steps in an attempt at a proof, rather than assumptions]
> >     (i)  Point (h) above does not, however, change the license of the XFree code, it
> > only changes the permissions on the combined code.  This is much like if I distirbute
> > a CD containing The Star Spangled Banner and music from Brittany Spears and Ricky
> > Martin.  I can only do it if I have permission of the copyright holders of all works
> > (even though one work may be in the public domain and the others not), and, moreover,
> > the restrictive license on the BS and RM works does not place a restrictive license
> > on The Star Spangled Banner, which remains in the public domain.
> >
>    The first sentence is false.

We've debated this above, and I'd like to see your responses to my comments there.

> Further, while your example is true, what
> you don't note is that the license on the BS can require a restrictive
> license on the CD as a whole (for that matter, a recording of a PD work
> still has its own copyright protection).

Yes, I do note that, read it again:  "I can only [distribute the collection of works] if I

have permission of the copyright holders of all works".

> You may, of course, obtain the
> individual works from a different source under a different license, in
> which case you'd have different rights with respect to that copy.
> Lynn



Reply to: