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

Re: Proposed statement wrt GNU FDL

Scripsit Anthony Towns <aj@azure.humbug.org.au>
> On Sun, Apr 20, 2003 at 05:35:14AM +0300, Richard Braakman wrote:

> > -- would you prefer that they hadn't seconded the
> > proposal either?  We could have had a nicely silent majority.

> I don't really see much value in "me too" posts. We build consensus by
> responding to criticism,

Me-too posts are a way of observing consensus. One rant (advocating a
possibly controversial course of action leading straight into major
flamewar territory) followed by complete silence does not establish a
consensus. It merely eastablishes apathy. A rant followed by a handful
of me-toos and still no criticism voiced allows consensus to be

Of course, a rant followed by constructive discussion about how to
implement the proposal in practise is even better, and once that ball
gets rolling, me-toos are superfluous.

> and there hasn't been *any* internal criticism of this stand since
> last November,

True, we knew we had a consensus that we don't like the GFDL and find
it lacking with respect to the DFSG if any of its options are
exercised. But we didn't previously have a clear consensus on turning
that feeling into action, and when the action is as momentous as that
lurking in horizon, I think Branden was right in not assuming that
consensus of assessment would automatically imply consensus on action.

We seem now to have the latter, which is good.

> There's been some question whether the front-cover texts are DFSG
> free. Considering we accept the obnoxious advertising clause, I can't
> see any reason for them not to be.

Perhaps the O.A.C. ought to be our next target, but let us fight one
battle at a time.

We do not accept the O.A.C. because we like it, but because of the
logistic problems of tracking down the well-meaning but misguided
authors who used it because it was in the template they were
following. That situation is unstable and marginally acceptable only
because there are probably none of those authors who actually care to
enforce it.

On the other hand, we have to assume that the FSF really mean what
they write in a brand new license, the drafts of which whey have
solicited and received extensive responses from the community.

> the hypothetical example of random people who don't have RMS's
> credibility attaching their own manifestos to free software
> documentation as some weird unerasable graffiti are both convincing
> to me. Are they convincing to anyone else?

While we should definitely include the hijacking example, some care
should be exercised in phrasing an explanation of what we think it
proves. In particular it should be very clear that we do not claim
that the possibility of hijacking in itself contributes to
DFSG-nonfreedom. (For example, BSD-licensed software and documention
can be hijacked too). On the other hand, the hijacking scenario does
help explain why we're mystified to see the FSF backing the license as
a "copyleft".

> The Problem
> ~~~~~~~~~~~

> The GNU FDL includes a number of conditions, which apply to all modified
> versions, that disallow modifications. In particular, these are:
[K and]
>  * L. Preserve all the Invariant Sections of the Document, unaltered in
>    their text and in their titles. Section numbers or the equivalent
>    are not considered part of the section titles.

> However, modifiability is a fundamental requirement of the Debian Free
> Software Guidelines, which state:

I think it would be good to emphasize more in the central section that
it is the stickyness rather than the rigidity of Invariant Sections
that bugs us. How about something like this after the DFSG quote:

  In particular, the DFSG requires that it must be legal to derive
  a new work by the particular modification that consists of removing
  one or more Invariant Sections. But this is expressly disallowed by
  condition L of the GFDL.

  If the rule had, instead been, that Invariant Sections could not
  themselves be modified, but could freely be omitted entirely in
  derived works, Debian would be able to distribute GDFL'ed
  documentation. If necessary, the Debian maintainer could himself
  remove the Invariant Sections, but in most cases where the license
  were not obviously abused or misapplied, we would still be able
  to include the original unmodified documentation based on a
  case-by-case review of the relevance and technical importance
  of the Invariant Section. (See the attached FAQ, question XYZ
  for discussion).

> Given the GNU Projects influence on Debian, shouldn't the GNU Manifesto
> be included in the Debian GNU/Linux Distribution anyway?

I propose expanding this question to:

   Why does Debian want to remove (say) the GNU Manifesto from the
   manuals? Do you want to suppress Stallman's views and just benefit
   from the software he created?

     The question is one we often hear, but its phrasing betrays a
     misunderstanding. Debian does not want to remove the GNU
     Manifesto from the packaged GNU manuals that we distribute.  On
     the contrary, we take pride in providing the best possible
     packing of the software we distribute. This includes making
     packages that are as close as possible to the upstream author's
     standard versions of the software, to the extent that this goal
     does not conflict with the practical utility and technical
     coherence of the Debian operating system.

     What we do want is for our *users* to be allowed to remove the
     GNU Manifesto from the manual if they can think of a reason to do
     so. In other questions we describe reasons that we imagine that
     users could think of, but even if we couldn't imagine them, we
     would not want to restrict our users' freedom by our own
     imagination.  We promise our users that if they have a reason
     that makes sense to *them* then they are always allowed to Make
     The Necessary Changes to the software they get for us.  That
     promise is part of the value Debian adds to the software we
     distribute.  It may look minor compared to the technical
     integration work that the majority of our developers' time is
     spent with, but it is actually at least as important for a
     significant segment of our users.

     If only we could be sure that the license on the manuals would
     allow a user who thinks that "because!" is reason enough for him,
     to remove the GNU Manifesto, we probably could still distribute
     the unmidified manuals with the Invariant Section in it. That
     would mean that part of what we distribute (namely the Invariant
     Section itself) would not, strictly speaking, be modifyable, but
     exceptions can be made for things that are both sufficiently
     non-software-like not to need modifyability for technical reasons
     and sufficienly relevant not to just constitute a waste of space
     in the distribution.  Of course both of these limits are
     judgement calls, and each particular Invariant-But-Removable
     section will have to be considered on a case-by-case basis.
     [Hmmm.. so I think at least, but I'm not sure that this is
     a clear d-l consensus. -HM]

     (For the record, the Debian Project neither officially agrees
     with the entirety of the GNU Manifesto nor officially disagrees
     with any particular part of it. But most Debian developers
     recognize individually that the ideas and ideals it seeks to
     express are to a very large degree the same ideas and ideals that
     lie at the heart of the project and are embodied in its Social

I suppose there should also be a QA-pair that addresses Georg Greve's
droits-moral argument, but I don't even understand it well enough to
write it down as a FAQ question. At least the Danish concept of droits
moral (which is the one I know) is not one I can see how to bend in
the direction of invariant sections at all.

Henning Makholm                        "Nej, hvor er vi altså heldige! Længe
                                      leve vor Buxgører Sansibar Bastelvel!"

Reply to: