Re: Final Draft: Interpretive Guideline regarding DFSG clause 3
-----BEGIN PGP SIGNED MESSAGE-----
On Wednesday 12 December 2001 10:04, Branden Robinson wrote:
> On Wed, Dec 12, 2001 at 08:15:23AM -0600, ichimunki wrote:
> > So perhaps this is a necessary move on the part of the Debian
> > community to get the FSF to clean up the emacs package a bit-- but it
> > might be better swept under the rug for now and submitted as a bug
> > patch.
> This has already been discussed with the FSF; please the list archives
> of debian-legal as referenced by URL in my proposal.
I meant the confusing license situation in their emacs tarball as a whole.
Their position on the manual, the Manifesto, and the freeness of the GFDL
itself is quite clear and apparently inflexible.
> I broached several alternatives with RMS, including one which would
> guarantee that at least one unmodified copy of all invariant texts
> within a package. He rejected all of my proposals.
How is he to prevent the addition of rebuttal essays or "clarifications" or
any other appended invariant sections related to existing invariant sections?
The GNU FDL seems to allow this. I don't care what he says, I don't find a
clause prohibiting writing such addenda (even as added invariant sections) in
> Please review my discussions with RMS on these very matters.
I read them the first time around. He seemed to be convinced that only an
idiot would try to abuse the GFDL (since a child would figure that out and
reject the manual outright)-- never mind that in the minds of some, he
appears to some to be abusing it himself.
> If you want to offer a counter-proposal that no piece of work that
> identifies itself as part of the GNU Project will ever be regarded as
> non-free, ever, please feel free to do so. I, however, will not support
I am not proposing that, I wouldn't propose that, and I wouldn't expect you
to! I merely think it would be the height of irony for any piece of GNU
software to be excluded from Debian GNU/Linux main.
> You're perfectly coherent; unfortunately, you don't seem to be recalling
> quite large portions of the discussion leading up to this proposal.
Could be. But I have read every message in the thread, but I'm sure I got
lost in the several very circular discussions. :)
> My previous proposal ("REVISED PROPOSAL") would in fact have let the GCC
> and Emacs manuals be interpreted as DFSG-free works. It was rejected by
> other developers on this list.
Was this because they find texts like the Manifesto completely unacceptable
in a package purporting to be free, or because they had concerns about using
a byte limit to determine the validity of the usage of invariant sections,
rather than some qualitative measure?
> In any event, my proposal does not forbid the grandfathering of any
> particular package in main. It also doesn't forbid making exceptions in
> the future. It's an interpretive guideline.
FWIW, I think your proposal is spot on. I agree that the GNU FDL is open to
abuse, even if I also think that no one in their right mind would attempt to
subvert it to distribute a novel.
I also have to say that the FSF, (and whoever else wants to write long-winded
licenses of any variety) can attach all the licenses they want to things.
That doesn't change the law (at least in the US). If I want to quote a line
or two from the emacs manual, indeed from the Manifesto itself, RMS can't
stop me-- even if I do it out of context and try to make him look bad! As
long as the quote is verbatim, that's called Fair Use. Note: I am not a
lawyer and this is not legal advice, just a statement of how I, myself will
act if I should so choose. Any lawyers reading this are invited to correct
any incorrect statements I might make regarding the law.
> DFSG 3 as I have is not going to be unobjectionable to most people. (In
> other words, who's going to be shocked that we think it's okay for
> copyright notices and license texts to not be modifiable?)
The question is, is the Manifesto an example of an exception granted by 3b of
your latest revised proposal? Is that essay's connection to emacs substantive
enough to consider to accept it as free? I ask the question somewhat
rhetorically-- it might be good to have some examples in 3b to use as a
benchmark for any future documents that need to be tested. But I'd also like
your opinion on the application of 3b for the Emacs manual, and whether you
think the Debian community (as you've been leading this discussion) agrees
with your interpretation of your own proposal.
> How consistent Debian wants to be with its stated principles in the
> future is not something I have sole power to determine.
Nor I. And to that end I think your proposal is very good, because it
prevents those of us who use only the main tree from being surprised by a
piece of fairly non-free documentation, when we thought we were getting only
free stuff. The hard part is saying exactly where does freedom start and end.
I don't buy the old canard "your freedom to swing your fist ends at my nose".
I'd say you better stop swinging long before that. And I'd be a lot freer if
you didn't raise your fist at all-- since that implies a threat of immediate
harm, hence has a coercive effect. So there is no clear standard. Similarly
there is some grey area as regards what kinds of invariant sections really
are free when the document carries the GNU FDL banner.
Obviously if the FSF is intending to lead by example, they think the
Manifesto is a good benchmark for what kind of texts should be allowable as
invariant sections. For my part I agree.
But if a sizable contingent of free software users, especially Debian
partisans, think otherwise, then I'd like to know what they think *would* be
acceptable under 3b in your proposal, if anything. Otherwise you may as well
strike the exception, since it leaves that grey area intact without providing
any guide to resolving questions.
And for my part I would prefer the Debian definition of
software/documentation freedom if it contradicted that of the FSF. It's not
up to the FSF to tell the rest of us what freedom is. They might be excellent
hackers and ground-breakers in the world of free software development, but
they didn't invent freedom, nor do they have any exclusive rights on it now.
The Debian definition is being arrived at through discussion and consensus,
not fiat-- and that makes it preferable in all cases.
--Michael Libby ( email@example.com )
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org
-----END PGP SIGNATURE-----