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

Re: The old DFSG-lemma again...

On Mon, Nov 05, 2001 at 08:44:34PM -0700, Richard Stallman wrote:
> Your message seems to start from the premise that the GNU FDL is
> unacceptable.  That is a rather controversial assertion, and you gave
> no grounds for it.

Sorry, this is the sort of thing that happens when I CC people out of
the blue.  (I should note that I personally tend to treat any new "free"
license with great suspicion, given my strong personal distate for
license proliferation, and in the interest of fairness I have to be as
critical of new FSF licenses as licenses from any other organization,
even if I trust the FSF in general far more than I do organizations
like, e.g., IBM or Apple.)

Specifically, I think the GNU FDL  -- or more precisely, some works
licensed under its terms -- could be interpreted to fail DFSG 3:

  Derived Works

  The license must allow modifications and derived works, and must allow
  them to be distributed under the same terms as the license of the
  original software.

Does the GNU FDL categorically fail the above?  Of course not.  Only
when exercised in certain ways, as you pointed out.  The problem I have
is that this places a greater burden on people in charge of vetting
licenses.  For instance, when a piece of software is submitted to Debian
and purports to be licensed under the GNU GPL, then -- in general -- it
is completely unproblematic.  It's safe to assume that the
author/submitter has not engaged in copyright, patent, or trademark
infringment.  Even if he has, these problems are severable from the GPL
itself.  With the GNU FDL, however, I am afraid it is possible for
authors to hamstring people's freedoms even within the bounds of the
license itself, by being excessive with regards to marking things as
Invariant Sections.  If the author is doing so illegitimately, so what?
The GNU FDL doesn't provide a "safe harbor" for people modifying
material licensed under it.

Quoting Section 1 of the GNU DFL:

  A "Secondary Section" is a named appendix or a front-matter section of
  the Document that deals exclusively with the relationship of the
  publishers or authors of the Document to the Document's overall subject
  (or to related matters) and contains nothing that could fall directly
  within that overall subject. (For example, if the Document is in part a
  textbook of mathematics, a Secondary Section may not explain any
  mathematics.) The relationship could be a matter of historical
  connection with the subject or with related matters, or of legal,
  commercial, philosophical, ethical or political position regarding them.

  The "Invariant Sections" are certain Secondary Sections whose titles are
  designated, as being those of Invariant Sections, in the notice that
  says that the Document is released under this License.

Well, what if some author violates the spirit of these definitions?  The
license itself grants no remedy for the users of the work.  Either the
work is de facto restrictively licensed under the GNU FDL (no
modification), or it is illegitimately licensed and falls into legal
limbo (unacceptable to the Debian Project) or reverts to unlicensed,
all-rights-reserved, heavily-protected-under-copyright-law status.

There is no such risk under the GNU GPL, or the OPL, as I recall.  With
these licenses I know I am free to modify and redistribute the entirety
of, or any part of, a work so licensed (except for modifying or removing
copyright notices and license text, of course).

> That is based on a misunderstanding of the GFDL.  The GFDL says that
> invariant sections must cover only topics of how the work relates to
> the authors or publishers.  So the shenanigans you describe are not
> actually permitted.

My point is that there is no effective means of prohibiting them.  Who's
going to stop the Author from doing so?  The License itself has no
provisions for rendering GNU FDL content free even if the Author gets
carried away with denoting parts as Invariant Sections.

> But suppose the GPL did permit them: what would that imply?  It would
> mean that the license was not a copyleft; non-free modified versions
> would be allowed.  That would be unfortunate, but it would not make
> the license any worse than the X11 license.

I disagree with your likening it to the non-copyleft X11 license; with
the X11 license, when a work comes to me *under that license* I know I
have uncontroversial permission from the author to modify the work.
With the GNU FDL, I don't have any such guarantee.  I have to scour the
work for Invariant Sections.

> If we imagine a a manual that misuses the GFDL by labeling technical
> material as "invariant", that would not be free, and Debian ought to
> reject that manual, as would the GNU Project.  But that is no reason
> to reject manuals in which the GFDL is used properly.

Would it be objectionable to the Free Software Foundation if the Debian
Project adopted a policy of accepting works licensed under the GNU FDL
as Free Software only if they contained no Invariant Sections or Cover
Texts (except for narrowly-construed notices of authorship, and the
license text)?  I hasten to add (again) that I am speaking only for
myself here, not for the Debian Project.

This would, ironically perhaps, make the GNU FDL very similar to the OPL
from Debian's perspective ("only DFSG-free if none of the license
options are exercised").  Given that possibility, can you address a
point that I perhaps left too implicit in my previous mail to you?  What
reasons would I, as a software author, or as a DFSG gatekeeper for the
Debian Project, have to prefer the GNU FDL over the OPL?  Is it
important to the Free Software Foundation that the GNU FDL be promoted
as an alternative to, or in preference to, the OPL?  If so, why?  Would
the Free Software Foundation be add to its roster of licenses something
morally equivalent to the license of the GNU CC Manual (at least as it
existed a couple of years ago -- it may have since been relicensed under
the GNU FDL)?  Such a license is completely unproblematic for Debian and
the DFSG as far as I can tell.

I should note that some of these issues arise -- I think -- because the
OPL predates the GNU FDL and has enjoyed some adoption before the GNU
FDL was finalized.  The GNU GPL and LGPL, however, have only within the
past few years started to see "competition" from licenses that attempt
to address some of the same concerns that the copyleft principle does,
unlike the MIT/X11 or BSD licenses.  And even so, the GNU GPL and LGPL
are clearly far more widely adopted than any of the other "Public
Licenses" that have recently come into view.

Thanks very much for your insights on these issues.

G. Branden Robinson                |
Debian GNU/Linux                   |     Music is the brandy of the damned.
branden@debian.org                 |     -- George Bernard Shaw
http://people.debian.org/~branden/ |

Attachment: pgpyhWcVcZsXn.pgp
Description: PGP signature

Reply to: