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

Re: The old DFSG-lemma again...



On Mon, Nov 05, 2001 at 12:10:36AM +0000, Sunnanvind wrote:
> 2) The GNU FDL doesn't pass the DFSG. This is a minor issue; FDL-stamped 
> documentation can reside safely in non-free. (But that's weird. And they 
> aren't there now. So we're obviously doing something wrong.)

[N.B.: it's a popular practice on the Debian lists these days to confuse
people's personal opinions with official Project dicta, so pay attention
-- these are my opinions only and I don't presume to speak for anyone
else.]

[RMS: sorry to CC you but I was unable to broach these issues with you
in person last week as I had hoped.  I'll understand if you do not have
time to reply.]

Or the FSF is.  I didn't get to go to Chicago last week to see RMS give
a presentation, so I didn't have the opportunity to discuss this subject
with him in broad strokes (that is, to determine the guiding principles
behind the GNU FDL as distinct from the GPL).

Indeed the OPL -- as long as no options are exercised -- strikes me as a
license much more congruent with what *I* understand to be the FSF's
principles than the FDL is.  Some people consider it quite a sport to
accuse RMS of one variety of ideological heterodoxy or another, but in
my assessment he has been strikingly consistent over the years.  So
it is likely there are motivations behind the GNU FDL of which I am
unaware.

That said, even if I understand the whys and wherefores behind the GNU
FDL, it doesn't follow that I'll think it's appropriate to Debian, at
least in its present form.  The limit of 100 on paper copies strikes me
as arbitrary and gratuitous, and more importantly, there's nothing
stopping an author from marking an entire work as invariant text, or
playing aggravating games such as marking only uninteresting text as
non-invariant.  Such a work would be unacceptable to Debian.

My personal stance is that I'm willing to grant immutability to license
texts themselves -- at least when tied to a specific work[1] --  but not
really anything else.  I find it difficult to think of an objective
criterion for what kinds of data qua data could be permitted to be
immutable, without inviting more arguments that you stand to resolve.
(Are hinted fonts data or software?  Images?  Music?  Philosophical
texts?  White papers on cryptography or compiler design?  Software
reference manuals?  Tutorials?  Technical documents liberally sprinkled
with asides, commentary, jokes, or even cartoons?  What about literate
programming languages?  What if I use haiku for documentation strings in
Emacs LISP, Python, or Perl's POD?)  For me, it's simple: immutable !=
free.  Most other stances, IMO, invite ambiguity and flamewars.

The bottom line for me is this: Debian's Social Contract focuses our
Project on software.  The packaging of anything that is not, or is not
easily regarded as, software, is of peripheral interest to us as a
Project.

[1] It's already copyright infringement to alter or remove copyright
notices, since respecting authorship is a pillar of the Open Source
philosophy, and not unimportant in the Free Software approach, I find
restriction on modification of copyright notices and license texts
acceptable and congruent with Debian's Social Contract.  On the other
hand, if we had a package that was a electronic book on software
licenses, containing the texts of many licenses, I would not accept the
package as free if the *contents* of such a text -- in contrast to the
license *on* the text, were held to be immutable.  Thus, the part of the
GNU GPL that says "Everyone is permitted to copy and distribute verbatim
copies of this license document, but changing it is not allowed" is
unproblematic when conjoined with material licensed under the GNU GPL,
but *is* a problem in my view when the GPL is regarded as a work in
itself.  I realize that this perspective may not be 100% compatible with
the FSF's feelings about the GNU GPL.  I think the FSF does mean for the
license text to be immutable any any form, because they don't want
"forked" versions of the license causing problems in the community;
however I might note that this horse appears to be out of the barn
despite their efforts.  License proliferation is in fact widely
perceived as a potential problem, and detrimental to synergistic forces
in Free Software.  Certainly the OSI's vigorous approval of various and
sundry licenses contributes to this problem; I guess it is their way of
justifying their existence.  That said, I am not sure the immutability
of any software license text qua text is really a practical problem at
the moment.

-- 
G. Branden Robinson                |
Debian GNU/Linux                   |           //     // //  /     /
branden@debian.org                 |           EI 'AANIIGOO 'AHOOT'E
http://people.debian.org/~branden/ |

Attachment: pgprsdEoY92h6.pgp
Description: PGP signature


Reply to: