Re: "object code" in the GPL and printed copies
Frank Küster <email@example.com> writes:
> Jeremy Hankins <firstname.lastname@example.org> wrote:
>> Typically that's the presumption (since object code is "not source"),
>> but that's really a question of law rather than the DFSG (i.e., get a
>> lawyer if it's important to you).
> It's important to me as a maintainer of a Debian package with some
> documents licensed with non-free licenses (GFDL, CC "Attribution
> non-commercial blabla") - naturally I'd rather persuade the authors to
> relicense their documentation than remove it. But I need a thorough
> understanding of the problems of documentation licensing to be able to
> do this. And of course I can't afford a lawyer, but an opinion of a
> couple of people more acquainted to law stuff than I am would maybe also
> help. Maybe I should ask the FSF...
Ah, I see.
>> As for the DFSG, I don't see how a
>> license that did not permit distribution of paper copies could be free.
>> Whether it's source or object code, it's still a version of the work,
>> and so the freedoms of the DFSG are still important (possibly with the
>> string of source distribution attached).
> I don't think that this is entirely clear.
Hrm. This sort of goes back to the whole debate about the meaning of
the term "software". Many (including myself, if it matters) think that
the most reasonable interpretation of "software" is as opposed to
hardware. This would imply that the paper the document is printed on is
not software, but the words on the page are. (Of course one could argue
that the printed words are analogue rather than digital, etc., ad
infinitum. And one could argue that a prohibition against printing is a
restriction placed on the bits on the HD as well. Thus are flame wars
In the end, of course, not everyone is going to agree. And frankly, I
think that's the more important issue. Many people who use debian are
going to assume that they don't suddenly lose the freedoms of the DFSG
just because a document has been printed on paper rather than recorded
on a CD. And if Debian is going to decide otherwise, it should be as
big and noise-making a decision as the whole bit about documentation
was -- perhaps even louder. Otherwise, I believe that we would be
violating our commitments to our users.
In so far as we're talking about the GPL, though, the software/hardware
issue doesn't enter into the discussion. I still think it's quite a
stretch to believe that the GPL gives one the right to restrict paper
> I have not yet made up my mind on this, so I'm just writing down some
> thoughts here:
> - Some people view intellectual property as "bad" (unethical, hindering
> development, or whatever) in any field, but other Free Software people
> do not claim this. For them, it's rather the particular field of
> "computer usage", with software, documentation, and possibly hardware
> that makes "Freeness" so important - be it for ethical or for
> practical reasons - while in other fields - like arts or literature -
> copyright is acceptable or even welcome.
> Since Debian and the DFSG are about Software, we cannot assume that
> everybody is of the first type.
I agree -- except that I would note that one could believe that free
licensing for stuff is good (even outside of the computer field) without
thinking that IP is necessarily bad.
> - Software and its documentation is "work in progress" most of the time.
> While an author might be willing to release a computer-related
> documentation to the public in its current form, this does not imply
> that they in fact think the work is fit for publicaton in form of a
> book, which is much more static, and which many people expect to be
> much better proof-read, typographically optimized etc. than is usual
> for a documentation PDF generated from texinfo, xml or the like.
> Furthermore, it's often not clear from the typeset text who is
> responsible for which content (only available in the CVS/SVN/... log
> and in source files), a point that might be deemed crucial by authors
> who have a reputation to loose. If I start contributing to a widely
> used documentation project *because* I find its present state
> inacceptable, I'd rather not see the intermediate product published as
> a book with my name (among others) on it...
> I doubt that we would be violating the spirit of the DFSG if we
> allowed authors to restrict printing because of such considerations.
Just because something is a legitimate goal doesn't mean that it's in
keeping with the spirit of the DFSG to write it into a license. Within
the free software world it is often expected that we use more
community-oriented mechanisms for achieving such ends (e.g., peer
pressure, polite requests, etc.) By using softer, more nuanced methods
(as opposed to the very hard and blunt tool of licensing) we permit more
freedoms and get more done.
>> Generally the concern with documents goes the other way: folks want to
>> make sure that paper copies can be distributed in classroom environments
>> and the like, where source distribution might be a significant
> That wouldn't be a problem with the GPL and a "written offer", and
> perhaps the concerns I raised are not important to people whom I need to
> persuade to switch away from the GFDL. But for people who chose some
> license forbidding commercial use (#345604, the license text in the
> initial bug report is outdated), it may be the major concern.
It can be a problem, actually. If a professor wants to distribute some
documentation for a class, it may not be straightforward to offer the
electronic version as well, and a 3-year commitment to honoring requests
for source could be a significant burden. This can generally be solved,
of course, but not with a straight GPL. The GPL along with extra
permission to distribute paper copies without source (perhaps under some
threshold number of copies) would generally do the trick, though.
I would also definitely recommend that you find out what issues people
have with relicensing rather than trying to predict them. Generally
things are neither as simple nor as complicated as we imagine them to
Jeremy Hankins <email@example.com>
PGP fingerprint: 748F 4D16 538E 75D6 8333 9E10 D212 B5ED 37D0 0A03