[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: LGPL module linked with a GPL lib

On 7/27/05, Jeff Licquia <licquia@debian.org> wrote:
> Excuse me for asking, but why is this monograph not freely available?
> Surely, as a non-lawyer, you have no hope of profiting from it, and
> having a succint, linkable statement of your arguments would do wonders
> for preventing such go-arounds as this one, and might even persuade
> laymen like me.

It's freely available by request, but not "published", for several
reasons.  One, I don't want to risk detracting significantly from
anyone's reputation in a way that isn't very firmly grounded in fact,
and so I prefer to gather as much criticism as I can privately before
presenting any conclusions in any form more official-looking than a
post to a random discussion list (where counter-arguments by other
people are only a click away).

Two, a couple of lawyers and law students of my acquaintance have
suggested (rather to my surprise) that it might be law journal
material, and academic journals like to be the first actual publisher
of a given paper.  (There's no profit for me in that; if law journals
operate anything like scientific journals, they'd probably charge me
for the privilege; but it's kind of cool to have a publication in a
peer-reviewed journal in any field, and if I ever want to go to law
school it can't hurt.)  Three, the paper (in LyX or PDF form) is
inappropriately large for distribution using a mailing list archive's
resources, and I haven't gotten around to resurrecting any of the
random places where I have hosted content in the past.  :-)

> "Use" is not my concern; distribution is.  Certainly, a user may combine
> Net-SNMP and OpenSSL in any way they want; even the GPL allows that.
> But (again, statically, so as to be entirely clear) when somebody puts
> Net-SNMP and OpenSSL together (or, to be more precise, OpenSSL and the
> Perl extension libraries) into a single binary and distributes the
> result, does that person have any copyright obligations to the Perl,
> Net-SNMP, and/or OpenSSL authors?

As far as I can tell (IANAL, TINLA), they have the same obligations
that they would have if they shipped printouts of the three projects'
source code in the same box.  Copyright is about creative expression
-- literal, human-readable text, images, and so forth, plus a limited
extension towards the non-literal that is historically applied to
translations to another human language, adaptations to another
dramatic form, sequels using the same characters and mise en scene,
etc.  It's not about interlocking functionality or about the idea
content of a work.

A copyright license is of course a creature of contract law, despite
the nonsense about "copyright-based licenses" that the FSF would have
you believe.  The GPL, in particular, is a perfectly valid offer of
bilateral contract, and as such can condition the grant of license on
whatever return performance is allowable in any other contract of
adhesion under the law in a given jurisdiction (in the US, mostly
state-level civil and commercial code).  But its strident claims to
divide the permitted from the forbidden exclusively using copyright
criteria do have some legal effect -- the offeree is on solid ground
when arguing that copyright law as judicially interpreted trumps
erroneous paraphrases in the GPL text and (IMHO) misrepresentations on
the FSF's web site.

> Furthermore, it's my sense that copyright holders have all kinds of
> discretionary power in restricting distribution of their works by
> others.  In what way is this power restricted when it comes to the
> dynamic linking question?

That discretionary power is channeled entirely through the terms of a
license agreement (as modified by statutory overrides such as the 17
USC 203 termination provisions), and power over "distribution" is
heavily curtailed by the doctrine of first sale.  That's not of any
great relevance to the GPL -- there are much stronger arguments
against the hypothetical dynamic-linking ban based on the GPL text
itself -- but it's useful background material.

> Of course, the FSF has had a relatively successful run at spreading
> their meme throughout the legal community.  The point was not that they
> are "pure" in some vague moral sense, but that they have
> well-established reputations and a long history of getting various
> portions of the legal community to see things their way.  Thus, looking
> at things from an "authority" perspective, you can see the force behind
> their arguments.

The legal community has been pretty resistant to this meme, as far as
I can tell.  Programmers and journalists are another story.  There's a
delicious romantic storyline about Robin Stallman Hood and Friar Eben
Tuck, and they have the advantage of what I think could fairly be
called a monomaniacal dedication to spreading that storyline over the
past 15 years or so.

Although I have no personal knowledge on the financial side, it
certainly looks to me like it has made them both rich men.  Little
snippets in the public record -- Jim Blandy's comment at
http://www.jwz.org/doc/lemacs.html about RMS's "luxurious pad on the
fourth floor of posh NE43", the indications from Moglen's letter to
Vidomi and Fluendo's defense of GStreamer that he has ways of
extracting revenues from his role in the FSF, back-of-the-envelope
calculations involving the typical conference speaker fee -- suggest
to me that their tax records would make interesting reading.  I could
be completely wrong; they could be scraping by on nominal salaries
from the FSF and Columbia University, plus a MacArthur fellowship here
and there; but it's enough for me to take their bizarre,
uncorroborated assertions about copyright law with a grain of salt.

> Except that I have already done this.  Arguably, I have not done it
> "enough".  But have you?  How would I, or anyone else, know?

Really?  Which citations from case law have you brought to bear? 
Presumably it will be settled in a courtroom someday, but the FSF
doesn't seem to need to have the law on its side when it has a large
number of decent, idealistic, vocal, IMHO naive programmers and
journalists plus a multi-million-dollar war chest from OSDL.

> Real legal degrees might be one way to vouch for the "completeness" of
> one's knowledge, but the FSF has those.  Prestige in the legal community
> might be another, but the FSF has that, too.  Success in negotiation?
> Also covered.  Success in the courtroom?  No one has that to any
> significant degree, and the lack of cases speaks to the effectiveness of
> negotiation, another point in the FSF's favor.

None of these carries much weight with me; but YMMV.  Law degrees are
obtained mostly by showing up and passing the tests; Moglen's
clerkship with Thurgood Marshall is a far more significant badge of
honor, but that was long before he started writing about "anarchism
triumphant".  I can't really speak about prestige, but I know several
lawyers (from me, this is pure hearsay, mind you; I'm totally not
qualified to judge, and I doubt any of them sees any profit in saying
so publicly) who are, let us say, more than skeptical about Mr.
Moglen.  Even if I knew to be true all of the things that I suspect, I
probably wouldn't have much of a case for attorney misconduct as such;
but if (as it appears to me) his harping on "copyright-based license"
has no basis in law, I would think it rather reprehensible in a
professor of legal history.  Attila the Hun and Napoleon Bonaparte
(not to invoke Godwin's Law or anything :-) had plenty of "success in
negotiation"; and where it matters (in the courtroom) the arguments in
the FSF FAQ haven't cut much ice.

> Perhaps we could get Eben Moglen to duke it out with you here.  If you
> can pull it off, more power to you.  Until then, what can I rely on?

I would doubtless be outclassed in a face-to-face (perhaps even a
mail-to-mail) with Professor Moglen; but it would be kind of amusing
to try.  In the meantime, I again commend to you the primary
literature at FindLaw.

- Michael

Reply to: