Re: RES: What makes software copyrightable anyway?
On 5/17/05, Raul Miller <moth.debian@gmail.com> wrote:
> On 5/17/05, Michael K. Edwards <m.k.edwards@gmail.com> wrote:
> > As a reminder: I favor the use of the non-crack-smoking GPL on a much
> > broader basis than it is used today. I would like to see all "open
> > source" projects, FSF and otherwise, relicensed exclusively under the
> > GPL so that we can roll up our sleeves and refactor the corpus without
> > worrying about whose chocolate winds up in whose peanut butter. I
> > believe that the FSF's bullshitting about copyright law is the main
> > obstacle to that happening in today's software world.
>
> I'm not convinced that the "crack smoking" argument is
> relevant, or accurate.
The "crack smoking" bit is hyperbolic, intended for at least somewhat
humorous effect, and more than a little self-mocking. Obviously IANAL
and lots of people, lawyers and otherwise, seem to buy the FSF's
interpretation. It seems blatantly obvious to me that the emperor has
no clothes, but I'm the voice crying out in the wilderness, "Make
straight the way!" (How's that for self-mocking hyperbolic literary
reference?)
> So far, my impression of that argument is that it rests very
> heavily on fine points of phrasing, but it also seems like
> you apply it rather liberally to anyone who draws different
> conclusions than you.
I apply it principally to the "GPL is not a[n offer of] contract"
assertion, and even so I usually limit it to the FSF's counsel, who
should know better. The argument against that claim doesn't rest on
any remotely fine points. Once that domino falls, I think the rest
follow quite naturally. It is certainly conceivable that a given
court of fact would rule that a given licensee's claim to have
accepted obligations only with respect to "derivative works" is
unpersuasive; but issues of fact are a crapshoot at the best of times.
> Your conclusions may or may not be correct -- ultimately,
> if that's resolved at all, that will be resolved by a judge.
> In the mean time... I don't really understand what points
> you are trying to make -- even to know if and when I am
> presenting something which is even relevant to an
> argument you are presenting. As a result, I don't
> understand what it is you're objecting to in the context
> of the FSF.
What is so hard about this? Would you like a summary with no
distracting references to case law?
The GPL is an offer of contract, no more and no less. Ambiguities in
contract terms are, by law in the US anyway, construed against the
offeror. GPL Section 0 defines a "work based on the Program" as a
"derivative work under copyright law", and that's the only candidate
for definition of that phrase that is still standing after
construction. Hence obligations to provide the source code for a
"work based on the Program" are limited to the scope of a derivative
work.
Under US law at least, a program that uses another program or library
through a stable published API is not a derivative work of the work it
uses. There is plenty of basis for construing blanket authorization
to create collective works into the GPL, and no basis for declaring
that distro CDs are OK but statically linked programs are not. A
fortiori, dynamically linked programs whose components are shipped in
separate packages, so that they are only combined at run-time, are
entirely exempt from claims of copyright infringement under 17 USC
117. So the Wicked Linker of Closed Programs to GPL Libraries (or GPL
Programs to OpenSSL) gets off scot-free.
There is a genuine concern with respect to the validity of the GPL,
which is that there is no mechanism in law (that I can see) to
implement the "automatically receives" language of Section 6 without
construing a grant of agency to sub-license along the chain of
licensee-redistributors. This particular kind of agency appears, at
least in some jurisdictions, not to be construable without quite
explicit language in a written contract. If this difficulty can be
resolved in a court of law, the "non-crack-smoking GPL" as construed
above is an excellent way for both businesses and hobbyists to do well
by doing good.
The FSF, for reasons known for certain only to its inner circle but
perhaps not unrelated to its business practices with regard to
embedded GCC/GDB vendors and other major donors, persists in
publishing a patently false set of claims about the legal import of
the GPL. The upshot of these claims is: "Free is Free and Closed is
Closed and Never the Twain Shall Meet (except when Free is used to
edit, compile, and debug Closed, in which case it's all good)." The
FSF also pursues its own mysterious agenda with regard to entities
like the Mozilla Foundation and the Apache Software Foundation that
have bent over backwards to cooperate with them as far as they can
without rendering the code of which they have custody unusable for
many commercial purposes.
Based on this understanding, and on recent history with regard to the
GFDL and other community relations fiascos, I believe that the FSF has
lost its moral and (if it ever had one) legal compass. With apologies
to Emile Zola, J'accuse...!
> It seems as if every time I've attempted to restate
> what seems to be your fundamental point, I've found
> out that I was incorrect. But, you are very emphatic
> about how anyone who doesn't draw the same
> conclusions as you -- be they courts, lawyers,
> or even merely myself -- has some very derogatory
> labels applied in those cases.
>
> Now, granted, there might be some chance that
> I and others are indeed smoking crack, even though
> I myself am unaware of this practice. But that's
> certainly not the only possibility.
>
> There is even the chance that you are
> incorrect about something or another (though,
> what that could be seems rather unclear,
> since I don't know how to test your assertions,
> nor, in many cases, what they are supposed
> to mean -- consider the "derivative and
> collective works are disjoint" as a fairly
> well discussed example).
Well, if that's your primary example, you seem to be correct in
stating that you don't know what it's supposed to mean. If I could
clear up your difficulty, I would; and it's not as though I (and
Humberto) haven't tried.
> > No, a license is an individual provision in a contract, whether
> > explicit or implied.
>
> Could you provide a citation on that? This seems to
> conflict with the usages I'm familiar with, including
> those in the urls you mentioned at
>
> > http://lists.debian.org/debian-legal/2005/01/msg00621.html
Let's see... A non-exclusive license is "a mere waiver of the right
to sue" (De Forest Radio Telephone v. United States), that's a
provision in a contract. It's significant continuing performance (in
re CFLC), that's a provision in a contract. An orally granted
exclusive license is modified under construction to a non-exclusive
one, and isn't retroactively terminated because "one party's breach
does not automatically cause recision [sic] of a bilateral contract"
(Jacob Maxwell v. Veeck). I can't find the conflict; what did you
have in mind?
> You might be correct here about what license terms
> are and are not, but even if you are I might be drawing
> the wrong conclusion from your phrasing. I'd be
> more comfortable if I had this from some official
> source.
How about Merriam-Webster's Dictionary of Law, as cited by FindLaw:
1 c: a grant by the holder of a copyright or patent to another of any
of the rights embodied in the copyright or patent short of an
assignment of all rights
And here's the most relevant definition of grant:
3 a: a transfer of property by deed or writing
b: the instrument by which such a transfer is made
You can chase the verbs "grant", "permit", "bestow", and "transfer"
around the dictionary until your face turns blue, but you're not going
to find anything more "official" than, say, Corbin on Contracts,
Nimmer on Copyright, and the cases they cite. If case law says (as it
does) that non-exclusive licenses need not be in writing but exclusive
licenses must, then waving around that "by writing" bit in the above
definition will get you exactly nowhere. Ditto your contention, which
is that there are people in the world who use "license" to mean
something other than a term in a contract.
Cheers,
- Michael
Reply to: