Re: LGPL module linked with a GPL lib
On 8/4/05, Raul Miller <moth.debian@gmail.com> wrote:
> Nevertheless, intellectual property is fundamentally different from
> real property, and the differences, in the general case, make it
> impossible to determine the boundaries of intellectual property.
It's a _little_ more abstract than real property ownership, which is a
lot more abstract than possession of a chattel; but it's rather less
abstract than, say, ownership of a 401(k) account -- a device where
you have limited control of some numbers in a brokerage firm's
computer, and the changes you request may or may not result in the
actual "trading" of some mutual fund "shares", which in turn once in a
while results in the "trading" of "common stock" of some companies,
which means God knows what. Yet the law has no great difficulty with
that kind of "property" either. If you're looking to set this kind of
a limit on the property abstraction, you're about 400 years too late
(joint stock companies; 300 for copyright, of course).
> Establishing property rights on intellectual property is like establishing
> property rights on shadows. Shadows very definitely exist, and you can
> write laws about them, but they shift and move depending on circumstances.
>
> And note that I've never claimed that intellectual property cannot be
> the subject of law. I've instead been claiming that such law can never
> equal the laws for real property. There's simply too much uncertainty
> about the domain for that to work.
Intellectual property law is so much simpler than real estate law (let
alone securities law), and so unlike your "property rights on shadows"
straw man, as to make this whole line of inquiry quite fruitless.
Copyright, patent, and trademark are imperfect but workable, and they
reflect a social consensus that may be fraying in spots (media
"consumers" do like their free beer, and the MPAA/RIAA counterattack
has also done quite a bit of damage) but isn't terribly amenable to
preaching from either direction.
> I said no one wants to take these issues to court, not that they are
> not taken to court. But I wasn't conveying my thoughts properly --
> what I meant to point out that letting the court resolve copyright
> issues is extremely unpopular -- it's just not wise in most circumstances.
Nobody wants to go to court if they don't think they have something
substantial to gain by it; going to court is expensive and risky. But
I certainly have no interest in letting anyone _other_ than
legislatures and courts resolve these issues if the parties directly
involved cannot -- not DRM schemers, not P2P "sharers", not captured
regulators, and not RMS either.
> In those cases, the problem is not one of physics, but one of
> choice. The properties could be surveyed more precisely,
> if someone cared enough to do so. With intellectual property,
> there is nothing to be surveyed.
Dude, it's just _not_that_hard_. The copyright, patent, and trademark
systems work just fine. How do I know? Because in almost every court
case in this area I read about, one side or both is a bunch of
assholes. Very, very rarely do you see the bellwether of bad law -- a
case where two parties who are minding their own business, coloring
within the lines, living and letting live, and peaceably making a buck
wind up duking it out in court. You get a few cases where both sides
seem sincere about thinking they were in the right, but wound up in a
competitive conflict that hinged on a subtle point -- cases like
Fogerty v. Fantasy and Lotus v. Borland. But you get a lot more cases
where it's hard to have much sympathy for either side, like Sun v.
Microsoft and Napster v. RIAA -- cases where greed meets greed and
they're either fighting over the scraps of a deal gone wrong (usually
an ill-conceived deal in the first place) or tussling over who
controls _access_ to a market where square deals are few and far
between.
> Compare:
>
> http://www.law.cornell.edu/copyright/cases/202_F3d_1227.htm
> "Aalmuhammed never had a written contract..."
>
> with
>
> http://www.gnu.org/prep/maintain/html_node/Copyright-Papers.html
> "Before incorporating significant changes, make sure that the person
> who wrote the changes has signed copyright papers and that the Free
> Software Foundation has received and signed them."
Sorry, I didn't make my point clear here, though I've made it on
debian-legal before. Aalmuhammed could not exert any of the rights of
a copyright holder because his contribution did not rise to the level
of authorship on any work that could stand in isolation. (Stand as a
coherent chunk of creative expression; please don't go haring off into
copyright-irrelevant engineering criteria like whether it can be
executed without linking to some other stuff first.) I do not find it
remotely plausible (IANAL) that patches, even quite extensive patches,
knowingly contributed on GPL terms and incorporated by a maintainer
who retains creative control, represent any obstacle to "GPL
enforcement".
RMS had a bad experience with Gosling Emacs; but that was a work of
authorship from which GNU Emacs was derived, not the other way around.
Unipress seems to have been content to let him expunge the remaining
literal Gosling text instead of pressing the point that the path he
followed would not have been possible without uses of Gosling's work
for which he could not demonstrate authorization; whether or not he
admits it, that was pretty damn decent of Gosling and Unipress. His
insistence on not accepting any contribution to a GNU work without
copyright assignment papers may stem partly from that and similar
experiences with other works he recycled, partly from a concern that
everyone take seriously a prohibition on borrowing snippets of third
parties' code, and partly from a mistaken belief that accepting
patches might result in a work of joint authorship.
But however sincere those intentions, the actual effect is to curtail
people's power to object if he changes the rules on them, as some
people feel he has done with the GFDL and as many fear he will with
GPL v3. With assignment papers, he's got meeting of the minds over a
set of written contract terms containing whatever covenants they
contain about what the FSF will or won't do in the future, and holding
him to any implicit covenants beyond the signed contract would be a
major uphill battle. Without them, he's got nothing but the "upgrade
clause" in a form contract and the fact that somebody e-mailed him a
patch, and I think it would be pretty hard to make that stick in the
face of concerted opposition by one or more non-trivial contributors.
Now as I said, I don't think I'd do anything differently in RMS's
position with regard to the XEmacs situation -- the mistakes made on
all sides there are long in the past, and it's best to just let the
fork go on its merry way. (And even if I thought I'd do it
differently, he is certainly within his rights to do it his way.) I
don't have any problem with the FSF's copyright assignment policy; I'd
even call it wise, given his natural inclination to want to retain
some power to act when some provision of the GPL runs afoul of the Law
of Unintended Consequences. Where he goes wrong, IMHO, is in not
fully respecting the law as it is, and in employing tactics on behalf
of his "ethically superior" strategy which he condemns when an
ordinary profit-seeking business tries them.
Cheers,
- Michael
(IANAL, TINLA)
Reply to: