Freeness of the GFDL
I'm basically a debian newbie (I'm not a debian developer or anything)
and (up to now) I've lurked on debian-legal because I find it
interesting (yeah, I'm weird).
Through all the talk of interpretive guidelines for the DFSG, etc, I've
been surprised at how little discussion has been given to the question
of whether or not the GFDL is, in fact, DFSG free. There seem to be
two, maybe three, schools of thought:
1) The freeness of the GFDL isn't terribly relevant. The DFSG is about
software, and however you draw the line between software & other stuff,
the stuff we're talking about here clearly isn't software. So we
don't really *care* whether it's free, whatever free might mean for
non-software data. Some people have floated the idea that we should
come up with other guidelines for data, others reject this approach. I
have to admit I don't fully understand this perspective, so I may be
misrepresenting it.
2) Sure, we care about whether or not docs are free (not just
distributable), but there's room for a fair amount of flexibility here.
Some data can be non-modifiable, and that doesn't make in non-DFSG free.
And so the GFDL is potentially subject to abuse, but if used _properly_
it's ok. The question is how do we decide when the use is proper?
3) Yes, docs should be DFSG free, and yes certain things (copyright
notices, terms & conditions, and possibly things like the history
section of a GFDL licensed document) can and should be non-modifiable.
But non-modifiable sections should be quite limited in scope, and thus
some things that might get appropriately (as per the GFDL) labeled as
invariant (e.g., philosophical or historical commentary on the work)
would make it non-DFSG free.
Most people involved in the, er, discussion seem to be going with (2),
or possibly (1). What surprises me is that only a couple of people
have come out with (3). Certainly, from my naive perspective, (3)
seems to be the obvious approach. I'm bothered by the idea that some
hypothetical font-selector program author wouldn't be able to pull a
couple of pages documenting keybindings from the emacs manual without
including massive (compared to the size of the font-selecting program)
amounts of invariant text. Is it really desirable that people be able
to include a "History of the Making of Foo, and Why I Did It" marked
invariant when they write foo? Should Debian be obliged to include it
if they do? Because if that's not proper usage of the GFDL, I don't see
where it's prohibited.
So while I don't want to add fuel to the flame, especially as
vituperative as this has gotten at times, I would like to hear why the
people that have no problem with invariant sections feel that they don't
run afoul of the DFSG. I'm assuming it's not *just* because it's coming
from the FSF, and because we've had emacs in main for ages.
Maybe it's me that's confused (that's quite possible), in which case I'd
appreciate it if someone could enlighten me. But it seems to me that
this issue hasn't really been resolved yet, and until it has been it's
premature to worry about compromises and guidelines and such.
--
Jeremy Hankins <nowan@nowan.org>
PGP fingerprint: 748F 4D16 538E 75D6 8333 9E10 D212 B5ED 37D0 0A03
Reply to: