A single unified license
Richard Stallman
converting documentation into software and vice versa. They call for
software and documentation to be a single pool of material with
compatible licenses.
That goal goes way beyond what I aimed for when writing our licenses.
I wrote the GNU GPL as a license for software, and wrote other
licenses for documentation. I did not try to make software licenses
cover documentation or vice versa. The needs in these two areas are
different, so the licenses are often different--and incompatible.
For instance, here is the license we used for most GNU manuals before
the GFDL:
Permission is granted to make and distribute verbatim copies of this
manual provided the copyright notice and this permission notice are
preserved on all copies.
Permission is granted to copy and distribute modified versions of this
manual under the conditions for verbatim copying, provided that the
entire resulting derived work is distributed under the terms of a
permission notice identical to this one.
Permission is granted to copy and distribute translations of this manual
into another language, under the above conditions for modified versions,
except that this permission notice may be stated in a translation
approved by the Foundation.
Nobody here would deny that this is a free documentation license. It
has the advantage of brevity, but in terms of merging the material
into a GPL-covered program, there's no difference between this license
and the GFDL. Both are incompatible with the GPL.
I designed the GPL to be license for free programs: it provides the
necessary freedoms for using programs as programs. If you want to
print the same material as a book, you can, but you run into practical
inconveniences--for instance, the requirement to distribute
machine-readable source code does not fit well with the way books are
published. The GPL can be used on documentation files, but it wasn't
designed for manuals.
Likewise, I designed the GFDL to be a license for free manuals: it
provides the necessary freedoms for using manuals as manuals, and
tried to attract commercial publishers to publish free manuals (though
none was interested in doing so at the time I wrote the GFDL). That
job was hard enough; I did not undertake to make it a suitable license
for software as well.
The goal that people are now proposing is much more ambitious:
essentially, to have a single license scheme that covers both software
and documentation. The benefit of this would be to combine two
separate information commons into a single larger one. This commons
would not include all free software, nor all free documentation; TeX
and Apache would not be part of it, nor would free manuals published
under the simple license above. The change would nonetheless be an
advance.
It is not a straightforward or easy job, however. To design a license
that is good for both free software and free documentation, and that
is close enough to today's GPL that it qualifies as a successor
version, may or may not be possible. In any case, it will take a lot
of thought.
I intend to make the effort some day, but first I have to finish GPL
version 3, which faces other difficult questions.
--
To UNSUBSCRIBE, email to debian-leg...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Thomas Bushnell, BSG
Debian, like everyone, would love a single unified license. But
that's not the problem per-se. The problem is that we want all the
licenses to be free by a single definition. That some of the licenses
will be incompatible with each other is a problem, but not one that
impacts freedom.
The problem with the GFDL from our standpoint is *not* just that most
such text cannot be incorporated into a GPL'd program. As you rightly
point out, that was true for the old GNU documentation licenses too.
The problem about incorporation what we are concerned with is that
GFDL-with-invariant-sections text cannot be incorporated into a GPL'd
program: the problem is that it cannot be added into *ANY* free
program whatsoever. This is *not* true for text from the old GNU
documentation license.
Thomas
David B Harris
Richard Stallman <r...@gnu.org> wrote:
> <large part of original message excluded because it's not relevant to
> my question>
>
> I intend to make the effort some day, but first I have to finish GPL
> version 3, which faces other difficult questions.
There have been some statements made by people who weren't part of the
FSF as to the nature of the GPL, version 3. Obviously nobody in their
right mind would take what are essentially rumours as fact, but it does
make one curious. Are there any working copies of the GPLv3 available
for perusal? If not, do you happen to have any idea as to how much time
will be given for community review?
Walter Landry
> I intend to make the effort some day, but first I have to finish GPL
> version 3, which faces other difficult questions.
I have recently come to the conclusion that making a unified license
is the only reasonable course left. There was much talk on this list
last year about a Debian Free Content License, which could be
converted into the GPL. However, the conversion has to happen both
ways, which basically makes them the same license.
The GPL is already being used for documentation. I think legitimizing
and encouraging that use should be a high priority for version 3. It
is a fairly simple exercise to change "program" to "work". As you
noted, one problem is to figure out how to regulate copying so that
people can print books without onerous conditions. One possible
solution is to add to GPL section 3) a fourth method for compliance
d) Accompany it with information as to how to obtain, for a charge
no more than the cost of physically performing source
distribution, corresponding source. (This alternative is allowed
only for noncommercial distribution)
This is only meant as a suggestion. It weakens copyleft a little, but
it does legitimize informal sharing.
You, of course, have your own priorities.
Regards,
Walter Landry
wla...@ucsd.edu
Anthony DeRobertis
> d) Accompany it with information as to how to obtain, for a charge
> no more than the cost of physically performing source
> distribution, corresponding source. (This alternative is allowed
> only for noncommercial distribution)
Looks quite a bit like (b).
Branden Robinson
> For instance, here is the license we used for most GNU manuals before
> the GFDL:
>
> Permission is granted to make and distribute verbatim copies of this
> manual provided the copyright notice and this permission notice are
> preserved on all copies.
>
> Permission is granted to copy and distribute modified versions of this
> manual under the conditions for verbatim copying, provided that the
> entire resulting derived work is distributed under the terms of a
> permission notice identical to this one.
>
> Permission is granted to copy and distribute translations of this manual
> into another language, under the above conditions for modified versions,
> except that this permission notice may be stated in a translation
> approved by the Foundation.
>
> Nobody here would deny that this is a free documentation license. It
> has the advantage of brevity, but in terms of merging the material
> into a GPL-covered program, there's no difference between this license
> and the GFDL. Both are incompatible with the GPL.
Can someone remind me how exactly the license above is incompatible with
the GNU GPL? Material under this license seems as miscible with a work
under the GNU GPL as materials under the 2- or 3-clause BSD licenses
are.
> I intend to make the effort some day, but first I have to finish GPL
> version 3, which faces other difficult questions.
Until that time, perhaps you could encourage authors to dual-license
their documentation under the GNU GPL and the GNU FDL.
Thanks for making the statement you have, though I must point out that
while your message offered some insight into the "Copying in Quantity"
section of the GNU FDL, and why an analogue to the GPL's requirement to
distribute source code has been weakened, it did not address why
Invariant Sections, Cover Texts, Acknowledgements, or Endorsements make
the GNU FDL more appealing to publishers than the GNU GPL is.
--
G. Branden Robinson | Psychology is really biology.
Debian GNU/Linux | Biology is really chemistry.
bra...@debian.org | Chemistry is really physics.
http://people.debian.org/~branden/ | Physics is really math.
Richard Braakman
I think the key difference is that (d) does not require the distributor
to make any kind of commitment. It would be acceptable to say "Here's
a Debian CD. You can find corresponding sources at http://debian.org/".
That's a lot easier than "Here's a Debian CD. And here's my solemn
promise to provide source CDs for this Debian version to anyone who
asks for the next three years. Please wait while I go buy a CD burner."
(Note that 2(c) is not an option here because Debian doesn't use 2(b).)
I think 2(d) reflects what people already do in practice. It may even
dispel some of the misinformation and misunderstanding surrounding the
GPL, by making it clear that this alternative is NOT allowed for
commercial distribution.
Richard Braakman
christophe barbe
> On Fri, 13 Jun 2003 18:02:56 -0400
> Richard Stallman <r...@gnu.org> wrote:
> > <large part of original message excluded because it's not relevant to
> > my question>
> >
> > I intend to make the effort some day, but first I have to finish GPL
> > version 3, which faces other difficult questions.
>
> There have been some statements made by people who weren't part of the
> FSF as to the nature of the GPL, version 3. Obviously nobody in their
> make one curious. Are there any working copies of the GPLv3 available
> for perusal? If not, do you happen to have any idea as to how much time
> will be given for community review?
I look forward to read a draft of the GPL v3, since Hans Reiser did
mention that the equivalent of 'Invariant Sections' would be added
in the forthcoming GPL v3.
See his post to the debian mailing-list:
http://lists.debian.org/debian-devel/2003/debian-devel-200304/msg01295.html
A quote:
"You'll note that ReiserFS anticipated the GNU GPL V3 by including
clauses that forbid removal of credits in its license, and for a long
time I have been telling Stallman that he needs to get V3 of the GPL out
the door."
Christophe
--
Christophe Barbé <christop...@ufies.org>
GnuPG FingerPrint: E0F6 FADF 2A5C F072 6AF8 F67A 8F45 2F1E D72C B41E
L'experience, c'est une connerie par jour mais jamais la même.
Thomas Bushnell, BSG
> Can someone remind me how exactly the license above is incompatible with
> the GNU GPL? Material under this license seems as miscible with a work
> under the GNU GPL as materials under the 2- or 3-clause BSD licenses
> are.
"Provided that the entire resulting derived work is distributed under
the terms of a permission notice identical to this one."
Thomas
Branden Robinson
> Branden Robinson <bra...@debian.org> writes:
>
> > Can someone remind me how exactly the license above is incompatible with
> > the GNU GPL? Material under this license seems as miscible with a work
> > under the GNU GPL as materials under the 2- or 3-clause BSD licenses
> > are.
>
> "Provided that the entire resulting derived work is distributed under
> the terms of a permission notice identical to this one."
Hmm, yup. "Entire resulting derived work".
It would have been nice to be able to "promote" this license to the GNU GPL.
I suppose there is still hope for the GNU FDL, though.
Hmm. Interestingly, one cannot publish a book that contains both the
entire GDB Manual from 2000 and the GNU Emacs Manual from that year or
this one, thanks to the latter's invariant sections. Combining them
into a single work for commercial sale would violate both licenses.
(The traditional GNU documentation that RMS quoted had no invariant
sections, and several GNU manuals did not have them until they were
added when they were relicensed under the GNU FDL.)
Unless one doesn't regard mere aggregation as derivation, that is...
--
G. Branden Robinson | Good judgement comes from
Debian GNU/Linux | experience; experience comes from
bra...@debian.org | bad judgement.
http://people.debian.org/~branden/ | -- Fred Brooks
J.D. Hood
documentation (i.e., for "content") is a good one. Perhaps
this could be done in GPL version 4. I would hope that in
extending it, the beauty of the current GPL is preserved.
What is beautiful about the GPL is that it grants the licensee
total freedom to do what s/he likes with the work, with a
single well understood limitation: s/he cannot distribute an
improved version of the work under more restrictive terms or
conditions than those of the GPL. This restriction is not
a burdensome restriction on anyone who wants to contribute to
the commons: it simply rules out a certain dangerous sort of
exploitative behavior: the Embrace and Extend(tm) strategy.
The GPL places minimal restrictions on the use of the content
itself.
That is more than can be said for the GFDL. The GFDL places
several different kinds of restrictions on the content itself,
none of which appears to be either necessary or sufficient for
the effective protection of software freedom. We are supposed
to accept these restrictions on the grounds that RMS doesn't
find them too onerous. That just isn't good enough, at least
for Debian's purposes.
So I hope that a future unified content license is modelled
on the current GPL, not on the GFDL.
--
Thomas Hood
________________________________________________________________________
Want to chat instantly with your online friends? Get the FREE Yahoo!
Messenger http://uk.messenger.yahoo.com/
Richard Stallman
Richard Stallman
the GNU GPL?
Each one is a copyleft. The GPL says the combined work must be under
the GPL. The simple license says the combined work must be under that
license. Both cannot be true at once.
Anthony DeRobertis
On Saturday, Jun 14, 2003, at 07:03 US/Eastern, Richard Braakman
> That's a lot easier than "Here's a Debian CD. And here's my solemn
> promise to provide source CDs for this Debian version to anyone who
> asks for the next three years. Please wait while I go buy a CD
> burner."
> (Note that 2(c) is not an option here because Debian doesn't use 2(b).)
Well, the CD burner thing doesn't really apply; either I got that
Debian CD by burning it, from a friend, or buying it. In the first
case, I already have the CD burner. In the second case and third case,
I should have source disks or a 2(b) offer. In the third case (and
maybe the second case), the first sale doctrine applies as well, so I
don't have to worry about the GPL.
However, I see your point here. For me giving a Debian CD to my friend,
I shouldn't have to either worry about doubling the number of CDs (and
download time) or 2(b). I agree that 2(c) needs to be improved in light
of the seldom use of 2(b).
May I suggest:
d) Accompany it with information you reasonably expect to remain
accurate for at least one year as to how to obtain, for a charge
no more than the cost of physically performing source distribution,
corresponding source. (This alternative is allowed only for
noncommercial distribution in quantities of under 120 per year)
The wording needs some help, of course. But the major changes:
1) You must reasonably expect the information to remain
accurate for one (or should this be 3?) years. In the
case of Debian, you'd use ftp.d.o or archive.d.o.
2) You may only do it 120 times or less per year. This
number seems high to me... Maybe it should be less. I
sure don't give out 10 Debian CDs a month.
One problem I still see with the wording is that the information I
could give could be:
"Visit <http://www.google.com/>, search for foomajig 0.13.2."
> I think 2(d) reflects what people already do in practice. It may even
> dispel some of the misinformation and misunderstanding surrounding the
> GPL, by making it clear that this alternative is NOT allowed for
> commercial distribution.
Hopefully, at least one of the works distributed by someone doing
commercial distribution without even reading the license will be
registered with the Copyright Office. That way, they can learn about
statutory damages under Title 17. I'd hope people doing commercial
distribution would be a little more cautious; and, of course, if
they're not reading the license, then adding in this clause isn't going
to help.
Nathanael Nerode
>GPL 3 is not at the stage to ask for public comments.
Rumor has it that it will contain loads of stuff which Debian considers
non-free. This is a *problem*.
The FDL public comment period resulted in *no* significant changes due
to the public comments.
RMS has declared that he has the final word on anything the FSF does,
and refuses to give out the names of anyone else involved [1].
Getting information about his reasons for decisions about the
controversial parts of the FDL has been like squeezing blood from a
stone [2]. (It took about a year to get the information that he
considered invariant sections acceptable because he classed them as
'packaging restrictions'.)
The FSF is set up as a charitable corporation, which means its board is
self-perpetuating. It is currently run as RMS's personal autocratic
fiefdom [3], and there's really no chance of that changing without
RMS's assent given a self-perpetuating board. (Unless he dies, of
course.)
Previously, developers were willing to assign their copyrights to the
FSF because they trusted RMS/the FSF to preserve the freedom of their
code.
After the FDL fiasco, I no longer trust him to do that. I am waiting
for a definitive legal opinion from the FSF on whether I can relicence
my *own* contributions under a more permissive license (such as the GPL
v.2). (I do not appear to get that right from my copyright
assignment papers.) If that right is absent, I will have to cancel my
current copyright assignment to the FSF, in favor of a disclaimer
(since putting my work in the public domain is far better than
allowing it to be made proprietary by the FSF).
--
If we can anticipate consistent behavior from the FSF, we will see the
following:
1. The GPL v.3 will be presented "for public comment". It will contain
unacceptable non-free provisions with no good explanation.
2. The comment period will contain lots of complaints about this.
3. The final version of the GPL v.3 will be released, with no change in
the non-free provisions, and no explanation as to why.
At this point, it will be necessary for free software developers to do
the following:
a) Discourage "version 2 or later" licensing in favor of "version 2"
licensing
b) Encourage the forking of all FSF-copyrighted projects
I would personally start a fork of GCC.
Forking and relicensing is a slow, tedious process. Accordingly, if the
FSF is planning to release a non-free GPL v.3, we want to start the
process as soon as possible; waiting simply gives them a head start at
subverting free software. On the other hand, if the GPL v.3 will be
just *fine*, we don't *want* to cause the trouble caused by forking and
relicensing.
What free software developers want are reassurances that the FSF is *not*
planning to cause this nightmare scenario by making a GPL v.3 which is
unacceptable to Debian. We have *no* such reassurances. The recent
past leads us to be very suspicious.
The "no information coming out" attitude from the FSF means, sadly, that
we must believe the worst: that the FSF is planning to subvert free
software with the GPL v.3, and that RMS is trying to hold off on
supplying information as long as possible so as to leave the opposition
scrambling to catch up, as it is with the FDL.
I will be starting the following projects:
1. Informational website with reasons to avoid the GNU FDL, and how to
do so.
2. List of free software developers who oppose the GNU FDL.
3. Project to create free (GPL) manuals for GCC, and eventually other
projects with FDL'ed manuals. (This is partly awaiting the FSF's legal
opinion on relicensing of one's own contributions, since if we definitely can,
I just need to collect the contributions of willing developers and then
fill the gaps.)
I'll need help with all of these. :-(
[1] Personal communication.
[2] Archives of debian-legal.
[3] In addition to the evidence above, all requests for licensing
changes on FSF projects must go through RMS personally, which can be
testified to by many GCC developers.
MJ Ray
> The FSF is set up as a charitable corporation, which means its board is
Please, pick one topic and stick with it. Do you really think that any
common form of accountability mechanism would have made any difference to
this situation? Do you really think that the FSF view on desirability
of free documentation being contained in free works has a direct impact
on their attitude to free software?
I'm not sure that people are opposed to the idea of FDL, but some
have problems accepting its current form. I do. I'm still hoping that
FSF will publish their detailed reasoning about why a GPL-incompatible
documentation licence that cannot be classed as free software is
necessary. I don't care about legacy publishers and don't understand
why these people who don't share our views are worth causing all these
problems, so there must be some other reasons. Mustn't there?
--
MJR/slef My Opinion Only and possibly not of any group I know.
http://mjr.towers.org.uk/ jabber://sl...@jabber.at
Creative copyleft computing services via http://www.ttllp.co.uk/
Thought: "Changeset algebra is really difficult."
Richard Stallman
mention that the equivalent of 'Invariant Sections' would be added
in the forthcoming GPL v3.
Reiser's statement was inaccurate. For GPL version 3 we are
considering requirements for preserving certain limited author
information in the source code, and making explicit that other
GPL-compatible licenses that are present on parts of the code cannot
be removed from the source, but nothing beyond that.
Nathanael Nerode
>Reiser's statement was inaccurate. For GPL version 3 we are
>considering requirements for preserving certain limited author
>information in the source code, and making explicit that other
>GPL-compatible licenses that are present on parts of the code cannot
>be removed from the source, but nothing beyond that.
*Heaves big sigh of relief*
OK, I'm happy. :-) Thanks for the reassurances.
--
Nathanael Nerode <neroden at gcc.gnu.org>
Don't use the GNU FDL for free documentation. See
<http://home.twcny.rr.com/nerode/neroden/fdl.html>
Branden Robinson
> I look forward to read a draft of the GPL v3, since Hans Reiser did
> mention that the equivalent of 'Invariant Sections' would be added
> in the forthcoming GPL v3.
>
> Reiser's statement was inaccurate. For GPL version 3 we are
> considering requirements for preserving certain limited author
> information in the source code, and making explicit that other
> GPL-compatible licenses that are present on parts of the code cannot
> be removed from the source, but nothing beyond that.
So you are not considering adding Invariant Sections, Acknowlegdements,
or Endorsements to GPLv3? That is encouraging, if true.
--
G. Branden Robinson | There's no trick to being a
Debian GNU/Linux | humorist when you have the whole
bra...@debian.org | government working for you.
http://people.debian.org/~branden/ | -- Will Rogers
Thomas Bushnell, BSG
> GPL 3 is not at the stage to ask for public comments.
That was one question. The other, and more important, question was:
"Do you happen to have any idea as to how much time will be given for
community review?"
Thomas
Richard Stallman
"Do you happen to have any idea as to how much time will be given for
community review?"
Please remember that this is not a cross examination; you are free
to ask questions, but how and whether I respond to them is my decision.
I don't remember seeing that question--or whether I noticed it. When
a message has a generally hostile tone, I generally skim it rather
than reading everything in it. If I saw it, perhaps I felt that I had
nothing much to say about the question, because I have not started
thinking about it.
On general principles I would not want to rush things. So I expect
there would be at least a couple of months.
Thomas Bushnell, BSG
> Please remember that this is not a cross examination; you are free
> to ask questions, but how and whether I respond to them is my decision.
Of course, but please also remember that if you completely ignore a
question, people will need to try and guess at your reason for doing
so. And, often as not, the guess is something like "he's embarrassed
at the question and has no good answer." If you don't want people to
make such an incorrect guess, then you'll need to fill in the correct
information yourself.
> I don't remember seeing that question--or whether I noticed it. When
> a message has a generally hostile tone, I generally skim it rather
> than reading everything in it.
This is a very poor strategy, and I would recommend you to stop it.
Many people have the same strategy, but it never works out well.
Indeed, you should be aware that many many people who use this
strategy do so by ignoring everything you post: and I'm sure you don't
want that! If we are to have everyone working together, it requires
at the very least the courtesy of listening carefully even to messages
that are very hard to hear.
Thomas
David B Harris
Richard Stallman <r...@gnu.org> wrote:
> That was one question. The other, and more important, question was:
>
> "Do you happen to have any idea as to how much time will be given for
> community review?"
>
> Please remember that this is not a cross examination; you are free
> to ask questions, but how and whether I respond to them is my decision.
>
> I don't remember seeing that question--or whether I noticed it. When
> a message has a generally hostile tone, I generally skim it rather
> than reading everything in it. If I saw it, perhaps I felt that I had
> nothing much to say about the question, because I have not started
> thinking about it.
>
> On general principles I would not want to rush things. So I expect
> there would be at least a couple of months.
I didn't intend the message to be hostile; rather, a neutral request for
information.
Thanks for responding, at any rate - that was indeed the question I was
most interested in.
Walter Landry
> As you
> noted, one problem is to figure out how to regulate copying so that
> people can print books without onerous conditions. One possible
> solution is to add to GPL section 3) a fourth method for compliance
>
> d) Accompany it with information as to how to obtain, for a charge
> no more than the cost of physically performing source
> distribution, corresponding source. (This alternative is allowed
> only for noncommercial distribution)
>
> Can you explain to me the difference that you have in mind?
Mostly what is mentioned here
http://lists.debian.org/debian-legal/2003/debian-legal-200306/msg00154.html
In particular, this allows a professor to print out copies and
distribute it to their class, without having to jump through any more
hoops. There was a long discussion of this sort of clause on
debian-legal. You can start looking at
http://lists.debian.org/debian-legal/2002/debian-legal-200206/msg00047.html