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

Re: GPL and linking (was: Urgently need GPL compatible libsnmp5-dev replacement :-()

On 5/10/05, Jakob Bohm <jbj@image.dk> wrote:
> On Mon, May 09, 2005 at 01:51:53AM -0700, Michael K. Edwards wrote:
> > Well, I have to admit it didn't occur to me that anyone might think my
> > opinion represented an opposite consensus.  More widely recognized
> > debian-legal personages such as Raul and Andrew seem to think I'm an
> > idiot or a troll, and to have no particular compunction about telling
> > d-d so.  (For the record, I am neither, thank you very much.)
> <mode="Fake Irish accent, pitiful">
> I was thinking of the about 800+ poor souls who never read d-l
> because they are too busy being DDs and doing useful work, and
> how those poor misguided souls might mistake Mr. Edwards for an
> official spokesman for the d-l consensus on what constitutes the
> proper way to package software for Debian without attracting
> legal trouble.
> </mode>

(chuckle)  Well, if they did follow my advice, they wouldn't be doing
a damn thing differently anyway, except they'd be a little more
careful about saying "socially, we oughtn't do X" instead of "legally,
we can't do X".

> I believe there are multiple ways in which the text of the GPL
> can potentially be interpreted in a manner which results in the
> same conclusion as the one reached by the GPL FAQ.

Right.  They're just not the ways that a court of competent
jurisdiction would construe it, as several people around here (way
more legally qualified than I) seem to agree.

> In this thread of discussion, Mr. Edwards has presented a
> specific interpretation of the GPL text, which 1. Does not
> result in that particular conclusion, and 2. renders large parts
> of the GPL text redundant and/or without force.

Which happens in contract construction _all_the_time_.  Which is why
competently drafted offers of contract have separability, integration,
etc. boilerplate.

> Taking that particular reading as a premise, Mr. Edwards then
> makes a number of rather strong statements regarding the honesty
> and/or intelligence of any statement or action making different
> presumptions regarding the proper interpretation of the GPL
> text.  It would simplify the discussion if Mr. Edwards would
> distinguish between his primary opinions regarding the
> interpretation and force of the GPL and various conclusions he
> makes on the assumption that his particular position would
> eventually prevail.

Where, exactly, did I make a strong statement about someone's honesty
that wasn't prefaced by "if this person knows better -- which, to my
limited knowledge of the law, they ought to"?  Or do more to insult
someone's intelligence than say "my level of pedantry must be
insulting people's intelligence by now"?  I have certainly attributed
willful ignorance to a couple of people, but that's kind of the
implication of my persistent dissenting interlocution anyway.

> I see these and similar issues in 3 related, but different,
> contexts:
> 1. Just in case (local) law does require permission, it is quite
> useful for that permission to be explicit.  Not through estoppel
> or similar, but as a clear and explicit permission capable of
> standing on its own in jurisdictions where it is not automatic.
> 2. By explicitly and specifically not prohibiting activities
> which law or industry practice would not allow a license to
> prohibit, the license offerer may protect himself from claims
> that broader parts of the license should be ruled invalid or
> otherwise being weakened by a court enforcing the requirement
> that a license cannot prohibit those activities.
> 3. By explicitly and specifically avoiding demands that could
> function as a means of monopoly abuse, an attempt to protect
> methods of operation, etc.  The license offerer may protect
> himself from having the license ruled invalid or severely
> weakened based on such grounds.

I agree with these things 100%.  Except that I (IANAL) think an IP
lawyer reviewing the GPL, in a capacity as the drafter's counsel at
any time after adoption of the 1976 Copyright Act, should have known
at the very least that the GPL was attempting to do things that
weren't legal in his jurisdiction.  And now that most nations are
signatory to the Berne Convention, I think it's quite clear that the
FSF FAQ is wildly misleading in any jurisdiction.  There may well be
barriers to fixing this other than the lawyer's honesty, but I confess
that I don't see them.  I am not qualified to judge whether the
lawyer's role in tweaking subsequent GPL variants to avoid a court
challenge is actionable, but I can't say that it attracts my
admiration.  And that's as far as I will go to challenge anyone's
honesty here.

> Returning now to your reference to industry practice, I may have
> seen industry practice where statically linked binaries are
> considered to be much more tightly bound than completely
> unrelated files residing on the same disk.  Or maybe I have not,
> I'd rather not say.

Well, I certainly have; but I've also seen practice to the contrary. 
It would have been very interesting to see Progress Software v. MySQL
litigated and appealed, so that a court of law would have an
opportunity to rule on the matter.

> > Nah.  A commented edition is a commented edition.  Which, I think,
> > might make a GPL program plus its GFDL documentation form a single
> > derivative work, and thus undistributable except by the FSF --
> > especially if one has, say, modified the program and documented the
> > change.  I say might -- but I personally am disinclined to chance it
> > just at the moment, although I reserve the right to change my mind.
> > (Hopefully that statement won't estop me from anything.  :-P )
> I was not saying that a statically linked program might be
> called a commented edition.  I was saying that a statically
> linked program might be considered to be just as invasive to the
> integrity of the original work as a commented edition is to the
> unaltered uncommented text, in the sense that both kinds of
> combined string of bits / letters could be seen as repeatedly
> interrupting the narrative flow of the original text to state
> something else of varying relevance to the point of
> interruption.

But the _work_ is the _expressive_ content, and the source code is a
far better place to look for evidence about that than the object code
is.  Putting my copy of "Shakespeare's Collected Works" through a
shredder, followed by a King James Bible, doesn't create a derivative
work of the two -- even though Shakespeare makes biblical references
passim.  However, constructing a joint concordance of the two
certainly does create a derivative work.  And if you want to find a
borderline case between creative and mechanical transformation, there
it is.

> > The state of a GNU/Linux system at run-time is sort of like like a
> > bunch of three-ring binders strewn around the surface of a desk, with
> > a bunch of reprints taken off the shelf and stuck in them or left
> > around loose.  Some binders (statically linked programs) contain
> > additional copies of (almost) all of the reprints mentioned in their
> > front matter; others (dynamically linked programs) just have
> > references to shared copies.  The first few chapters contain the theme
> > of the binder, which might be in a single language (a binary
> > executable) or several (a data file stacked on top of an interpreted
> > program stacked on top of the binary that executes it).  The CPU is a
> > djinn that lives in the magic lamp labeled "Linux kernel" sitting on
> > the desk.  ;-)
> >
> > As I see it, there are "work" boundaries everywhere there's a work of
> > authorship that makes sense by itself, although you have to understand
> > the ideas in many others to read it.  In order to make mechanical use
> > of them the djinn in the desk has to follow the syntax back and forth
> > from work to work.  But that's completely irrelevant to any
> > copyright-based analysis.  Copyright is about concrete forms of
> > expression (and to a lesser extent about characters, settings, and
> > plot arcs), not about ideas or engineering relationships.
> An interesting and somewhat entertaining analysis.  Actually, I
> might (or might not) be inclined to considered the CPU etc. to
> be more similar to the mechanisms of a theatrical movie
> projector or a home TV or radio receiver.
> In this analogue, while the mechanics of those devices is of
> limited legal significance to the works they process, their
> usual and expected behavior as apparently presumed by the
> creator of the works or other collections of bits provided to
> them may (or may not) have a profound influence on the legal
> status of those works or collections of bits.

Right.  See Walt Disney v. Bourne (
http://laws.findlaw.com/2nd/947793.html ) for a brilliant and closely
argued example.

> An even simpler, but more distant, parallel could (or could not)
> be an ordinary printed and hardbound book, both in its final
> form and as the unfolded printed sheets of paper where the pages
> have been printed in a weird n-up layout which puts the pages in
> the "correct" order after folding and binding.  While the exotic
> mechanics of the bookprinting and bookbinding process is of
> limited legal significance, the creators or publishers presumed
> expectation that the text will eventually end up in the final
> book form and be read in the sequence and manner traditional to
> our western culture may or may not be of great importance when a
> court considers the legal status of the unfolded prints of a
> book whose legality is being disputed.

Oh, that's clever!  And honestly, I'd say that even if it didn't
support the argument that I'm making.  But it does, because the
grounds on which computer programs are afforded copyright protection
is their expressive content, as readable in their source code, and
that hasn't been hashed together.  That's not quite true for some
things that have expressive content at run-time, such as video games;
but it does in all the cases I care about.

> > > I argue no such third scenario.  I argue that the paragraph
> > > might not disclaim all attempts to reach outside the limits of a
> > > "work based on the Program", this hinges on whether "a whole
> > > which is a work based on the Program" covers a different scope
> > > than "a work based on the Program".
> >
> > I can't imagine reading the two substantially differently, given the
> > use of "which is"; but YMMV.
> If a legal document suddenly varies its language compared to the
> rest of the document, I tend to think there might be a legal
> significant in that.

Sure.  Except the natural way to look at that example is that "a
whole" binds with the phrasing "a part of a whole".  Yes, if you
diagram the sentence, there's a single line you can snip and you get
the noun phrase "a whole which is a work based on the Program"; but
it's not like some places it says "a part of a whole which is a work
..." and other places it says "a part of a work ...".  That's the only
usage of "part of" anywhere in the GPL.  So I think you'd have to find
something more explicit even if the GPL were a statute and you were in
principle obliged to find meaning in every variation of wording and

> I have heard of jurisdictions where simply saying "to the
> maximum extent permitted by law" etc. does not protect a clause
> from being ruled invalid.  But this is not what the GPL or its
> use does here.  The GPL explicitly limits itself from having
> specific legal effects even where those legal effects might be
> permitted by law.  As such a court might or might not rule that
> the GPL is not even trying to exceed the applicable limit of the
> law and rule much more favorably than if it had tried to demand
> those things and then added a "to the extend permitted by law"
> or "separability" cop out.

Sure.  But it remains an explanatory aside, clarifying the drafter's
idea of what the language above is trying to achieve (which it in fact
couldn't achieve without demanding exclusive license on derivative
works anyway -- and courts are very reluctant to construe that without
explicit language, and may be banned from so doing in some
jurisdictions).  It doesn't suddenly make "work based on the Program"
mean something else back in Section 0.

> > That's not my argument at all.  I am of the opinion that it reaches to
> > the boundary of the enclosing "derivative work under copyright law",
> > which is (for any given work under discussion) completely unambiguous.
> >  There may also be one or more enclosing scopes, each of which is
> > itself a work of authorship identifiable separately from its context.
> > Each of these is not a derivative work but a collection, either
> > copyrightable or uncopyrightable depending on the non-triviality of
> > its selection criterion.  (Geez, this is starting to sound like I'm
> > talking through a mouthful of mush.)
> I don't think we live in a world of concentric crystal spheres
> guiding the paths of each planet around the Sun, but then I am
> not an astronomer either.

Actually, I'm a lot more of an astronomer than I am a lawyer; the only
postgrad program I've attended was astronomy.  :-)  But that really
does appear to be the way that it works under the law.

> > That was not the intent of this analogy.  As I wrote, the MP3 player
> > vendor might have some legal means of retaliation.  But claiming that
> > the retailer has "anthologized" the MP3 player doco with the
> > dismantling instructions, or claiming that references to figures in
> > the doco make the instructions a "derivative work" of the doco, and
> > then trying to proceed directly to preliminary injunction under
> > copyright law just isn't going to get them anywhere.
> Since the MP3 manufacturer in your hypothetical is trying to
> protect the device, not the doco, I find it difficult to get a
> good parallel without treating it metaphorically as if the MP3
> player itself was a copyrightable work rather than a piece of
> hardware.  For a better variant of the analogy consider that the
> item being salvaged is not a CF chip but a pre-burned ROM
> containing copyrighted music not licensed for use outside the
> context of the player.

But my whole point is that to be a copyrightable collection you've got
to be combining two works on creative grounds, not unrelated grounds
of utility; and even then, you're creating a collection, not a
derivative work.

> > Er, that would be "a whole which contains a work based on the
> > Program".  "is" \neq "contains".  And it's just plain perverse to
> > claim that "a whole which is a work based on the Program" can be
> > interpreted without reference to the Section 0 definition of "work
> > based on the Program".  Nice try, though.  :-)
> I am not trying to interpret it without reference to the Section
> 0 definition.  I am considering the possibility of maybe
> interpreting the longer and different phrase as referring to
> different set of items implicitly defined *in terms of* the
> definition of "a work based on the Program" as defined in
> Section 0.  Put another way I am considering the potential
> interpretations of the prefix "a whole which is" when used in
> front of that phrase.

And "is" remains a verb implying equality.  This phrase means:
    choose "whole" \element W ( = \{ P \} \union D ).

> > > And does "a work based on the Program" indeed have the limited
> > > meaning debated elsewhere in this message?
> >
> > Ask a court.  I've beaten that horse as much as I'm going to.
> Which is why I was not arguing it here, just declaring a logic
> dependency on the (as yet disputed) outcome of that animal
> abuse.

Ah, OK.

> > > As I wrote, if the answers to these two questions are YES, YES,
> > > then the GPL looses its reach, and several of its permissions
> > > (including the "mere aggregation" paragraph) become no-ops
> > > granting permissions already granted.  But if either of these
> > > two questions are answered with a NO, then this paragraph is
> > > disclaiming the GPL's reach in fewer cases and the other
> > > paragraphs are no longer no-ops.
> >
> > Actually, there is a sense in which they're not no-ops.  I think they
> > estop the offeror from claiming in a courtroom that they didn't intend
> > to authorize the conduct that is explicitly permitted in them.  But
> > the contract's meaning is otherwise the same with or without them.
> > (IANAL, and all that.)

Note: my comment immediately above is incorrect.  Humberto has made a
better case that they're grounds for construing license to create, and
agency to license the creation of, collective works.

> In the "part of the operating system" paragraph there is
> explicit language making that exception conditional on said part
> of the operating system not accompanying the GPL work.  I have
> heard rumors that when Sun, Microsoft and others occasionally
> ship GPL works, they do so on a physically separate tape, CD or
> DVD to comply with that condition.  Similarly, in the OpenSSL
> scenario (see the subject line of this debate), a work shipped
> on CD#1 along with the Linux Kernel, gcc, glibc, OpenSSL and
> other works is normally considered incapable of taking advantage
> of that clause, but a GPL work shipped outside Debian but
> created to run on a Debian system where OpenSSL accompanies
> those major components of the Debian operating system might be
> able to link to the CD#1 copy of OpenSSL because of that clause.

That "part of the operating system" bit is another example of
reassurance about the drafter's intentions, in this case reassurance
that you don't have to include Sun's header files or C library (in .a
form) or something when you distribute a binary built with Sun's
compiler.  You're right that it isn't trying to encourage _vendors_ to
put GNU and UNIX on the same tape; I was thinking of system backups.

It's nice that they reassure me that way, but as far as they and I are
concerned they actually already OK'ed that when they defined "work
based on the Program".  Sun, on the other hand, might have something
to say about my distributing their C library, but on the grounds that
I've copied and distributed, not derived or collected.  And the bits
of their header files that are needed to create an interoperating
binary aren't copyrightable anyway.  Note that this is an area where
static vs. dynamic linking matters, not to the GPL but to a
hypothetical complaint from Sun; they could nail me if I linked
statically and distributed the result without license, because I don't
_have_ to copy that much of their work to interoperate with their
system.  Even if a dynamically linked C library wasn't available, I
could in principle write one.  Compare the Nintendo and Lexmark cases.

> I was thinking of cases where Y brings the GPL text to the table
> and tells X to make the GPL offer with respect to P in order to
> attain some benefit.

In a court case between your X and some licensee U, it's against X
that the offer must be construed.  Y has little to say about it.  If Y
believes X isn't adequately enforcing license on U, then Y might be
able to sue X for violation of a covenant to attempt to enforce; but
that's a totally different story.

> > It appears to me that Section 2b would impose such conditions if it
> > were not for the "a whole which is a work based on the Program"
> > language in the paragraph following 2c.  The remainder of Section 2
> > muddies the water further:
> >
> > 2b) ... any work that you distribute or publish, that in whole or in
> > part contains or is derived from the Program or any part thereof ...
> >
> > (paragraph after 2c) ... But when you distribute the same sections as
> > part of a whole which is a work based on the Program ...
> >
> > (next paragraph) ... the intent is to exercise the right to control
> > the distribution of derivative or collective works based on the
> > Program ... [the only place where the drafter mentions "collections"
> > or "collective works"]
> Could the textual proximity and relationship of those two
> paragraphs be read so as to define the "a whole which is" prefix
> to mean "optionally collective work of", given that there might
> be no other definition of that prefix in sight?


> > ... and then the "mere aggregation" clause.
> >
> > See why there's a principle of construction against the offeror?  The
> > only way I see to construe a single interpretation out of this muddle
> > is to say that the broad language of 2b is narrowed back to the same
> > scope as the rest of the contract by a restatement that only mentions
> > two cases -- distribution "as separate works" and  "as part of a whole
> > which is a work based on the Program".  The claim that "the intent is
> > to exercise the right to control the distribution of derivative or
> > collective works based on the Program", and the "mere aggregation"
> > clause, then reduce to estoppel fodder.
> Different legal systems might not go so far as to constructing a
> contract in a manner which would make it so much more
> inconsistent than another construction more consistent with a
> plain reading of the totality of its text.

The drafter of this offer of contract has shouted from the rooftops
that he has made no attempt to conform it with the applicable law. 
What is so surprising about finding it to be full of internal
inconsistencies when read under that law?  Perhaps you are thinking of
the principle of construing a _statute_ consistently with a plain
reading of it, which can indeed be at odds sometimes with the
"construe so that every word has meaning" principle.  That's one way
that judges wind up having to effectively rewrite statutes.

> > I repeat, that's not why it was denied.  Check out the full opinion at
> > http://pacer.mad.uscourts.gov/dc/opinions/saris/pdf/progress%20software.pdf
> > .  There are a standardized sequence of criteria for preliminary
> > injunction (at least under the US version of common law; YMMV), and
> > the judge followed that sequence to the letter.  MySQL would have to
> > have met all of the major criteria: likelihood of success on the
> > merits, potential for irreparable harm, and balance of equities/harms
> > (courts of fact rarely go into the effect on the public interest in a
> > routine contract law case).  They didn't meet any of them.
> >
> > The judge's stance on the "derivative work" issue (note: she appears
> > to have ignored the paraphrase and concluded that the factual dispute
> > hinged entirely on whether Gemini had created a derivative work) was
> > indeed to lean towards MySQL's interpretation, based on a review of
> > the parties' affidavits followed by oral arguments of which I have yet
> > to find a transcript.  But she then systematically trashed MySQL's
> > case that winning that point would meet the "likelihood of success on
> > the merits" criterion, let alone the others.  Had there been any
> > reasonable prospect of piercing the contract and proceeding to
> > copyright law standards, MySQL would have been entitled to an
> > automatic presumption of irreparable harm and a bye on the "balance of
> > harms" standard.
> What I said, the judge was not convinced that the likelihood of
> success on the merits was sufficiently high so as to grant a
> *preliminary* injunction.  This is very different than saying
> that there was a greater likelihood of the opposite outcome.

Did you read the MySQL opinion in full?  What possible further remedy
on GPL grounds can you imagine Judge Saris granting to MySQL had the
case gone to trial?

> And then there was the apparent ruling that the defendants
> corrective actions reduced the harm significantly compared to a
> hypothetical case where the defendant had taken no such actions.

Right.  So the facts of the case met _none_ of the three criteria:
likelihood of success on the merits, potential for irreparable harm,
and balance of harms.

> > I think it's way too strong to say that Judge Saris declared anything
> > of the kind.  I think (although I have not read their affidavit) that
> > Progress Software did not dispute acceptance of the GPL, and argued
> > exclusively that Gemini was not a derivative work of MySQL's GPL code
> > and that they had in any case acceded to MySQL's GPL-related demands
> > by the time of the hearing.
> OK, still I see no ruling against the GPL here.  No ruling
> against derivative work.  Only a failure of the injunction
> tactic in a case where the defendant had already corrected their
> behavior.

It was a quite firm ruling against the "GPL is a creature of copyright
law" theory, against the premise that it "automatically terminates"
(on non-material breach), and against the premise that relief would be
granted on a technicality in the absence of some demonstration of
actual harm.  Judge Saris could freely stipulate the 'linking creates
a work based on the Program' premise and still have no grounds for
ordering any remedy, either at the preliminary injunction stage or at

> I did not say they tried to avoid conflicts in general.  I said
> that the FSF seems to be trying to steer clear of some losing
> battles in order to preserve the strength of their licenses for
> more winable battles.

See my above comments on your "3 points".

> > Happily, contrary to what many people think, that's not how binding
> > legal precedent is created (in the US, anyway).  The only court that
> > can set a precedent with regard to a point of law is an appellate
> > court, and appellate courts have no power (in principle) to make or
> > change determinations of fact regarding the parties' conduct.  But
> > you're right -- since the FSF is engaged in social engineering rather
> > in clearing up people's confusion about the law, I fear they are only
> > interested in getting into a courtroom if they can engineer an
> > apparent "win".  I invite you to read the MySQL opinion side by side
> > with the FSF's press release (
> > http://mailman.fsfeurope.org/pipermail/press-release/2002q1/000035.html
> > ) and to form your own judgment about the motivations of the
> > individuals involved.
> I am more concerned about social engineering directed at the
> courts.  How many appellate court rulings tend to involve
> multiple points of law?  And how many of those rulings become
> precedents for cases where only some of those points of law are
> present?

Both are true of practically every case that is accepted by an appeals
court.  And certainly of every case that I have cited on debian-legal.

> > In my eyes, it's a catastrophic precedent, from which nothing but
> > instant out-of-court settlement could have distracted attention.  The
> > popular press seems to have taken the press releases at face value at
> > the time -- but then who has time to research when you're on a
> > deadline?  And it's so, so sad that the FSF has put so much of its
> > credibility into a set of claims about the legal significance of the
> > GPL that I believe will sooner or later be demolished in court in a
> > way that no press release can spin.
> You have yet to quote anything from that precedent that actually
> ruled against the GPL in any way.

Against the GPL as a valid offer of contract?  No.  Against the FSF's
interpretation?  See above.

> > Or, of course, if they do know better and they value their credibility
> > as individuals or as an institution.  They may have a bunch of law up
> > their sleeve that I can't find, or I could just be sadly incapable of
> > legal reasoning.  But I call your attention to the fact that an
> > attorney at law who is admitted to the bar in a given jurisdiction is
> > in fact an officer of the law, and is sworn to uphold the law to the
> > best of his ability, modulo a few very limited outs related to
> > attorney-client privilege.  If Eben Moglen has no reasonable basis for
> > a belief that his public statements about the legal implications of
> > the GPL are true (which I am not qualified to assess), then he is
> > skating on some rather thin ice.
> This is what I meant by "makes no legal difference".  If they
> believe that your analysis is correct, then it would make no
> legal difference for them to admit so and move on.  If they have
> any doubts (or even a firm belief to the contrary) then making
> such a declaration should be done only if it is in their
> interest to do so for reasons other than your analysis.

Those words "no reasonable basis for a belief" have special meaning in
assessing the conduct of an attorney-at-law.  As I said, I am not
qualified to assess whether they apply in this case; but if they do,
it could affect Eben Moglen rather badly.  And my comments on the
credibility of the FSF as an institution still apply.

Neither Mr. Moglen nor the FSF is under any obligation to respond to
such criticisms, of course; but I betray no secrets in saying that
their response to me to date has been rather unimpressive as evidence
that they have a reasonable legal basis for their public assertions.

> > > It is in Linus interest to assure the world that running on top
> > > of a Linux kernel does not subject your code to the terms of the
> > > GPL, and to do so in a manner seen as legally binding by those
> > > who might otherwise refuse to use his big creation.
> >
> > I believe that Linus chose the GPL at a date when the LGPL was also an
> > option, and did so with no intention of suggesting that userspace
> > programs would be obliged to be GPLed.  He's long since estopped out
> > the yin-yang from making any such claim, as is anyone with enough
> > knowledge of what a kernel is to submit a patch to the LKML.  Sneering
> > at Linus's "big creation" does not become you.
> I am not sneering at it, it really is a big creation, perhaps
> the greatest thing he will ever do to benefit mankind (but he is
> still young, and might still live to do even greater things). I
> am saying that to further this great benefit to mankind it is in
> Linus' interest to assure the world that there is no legal risk
> in running non-GPL software or drivers on a Linux-based
> operating system, even at the cost of imposing extra legal
> restrictions upon himself.

I apologize for misinterpreting what you wrote.  You are quite right.

> > > The FSF has a different interest in many cases, especially for
> > > its various GPL-ed libraries and it is thus in the interest of
> > > the FSF to stick to its GPL interpretation if at all possible,
> > > and grant explicit additional permissions in the cases where its
> > > software would otherwise not be used, or where the circumstances
> > > could otherwise inspire the creation of case law to the
> > > contrary.
> >
> > Well, I'll just say that gerrymandering the licenses on individual
> > software components to avoid exposure of a fraudulent legal
> > interpretation -- if that is in fact what the FSF is doing, which I am
> > not qualified to judge and emphatically do not assert -- does not
> > strike me as a way to retain the public's trust as the guardian of
> > copyright on a substantial corpus of software, much of it donated on
> > the basis of that trust.
> I see it more as culling the licenses where they could be ruled
> invalid for legal reasons not applicable to the license in
> general.  Like cutting a branch off a tree in your garden to
> prevent that particular branch from growing out on a public road
> where it could be declared a menace to public safety, causing
> the entire tree to be cut down at court order.

And were the tree not rotten to the core, I would agree.  But spraying
green paint on its leaves to fool the inspector when he drives by,
while knowing full well that it will smash the house across the street
in the next windstorm, is a different story.

> > Whew.  All of a sudden we're both getting very, very careful about our
> > disclaimers.  Disclaimers about disclaimers, no less.  In the course
> > of writing this message I learned for the first time about the
> > existence of Daniel Wallace's suit.
> The kind of suit that my 3 point list above is about.

The kind of suit, apparently, brought by a person who also does his
own legal research:
Curiouser and curiouser.  But I still think the Clayton Act claim is nutso.

> > To whom it may concern: my debian-legal and debian-devel postings on
> > the topic of the GPL are Copyright (C) Michael K. Edwards 2004, 2005
> > (as applicable), All Rights Reserved, and you plagiarize them at your
> > peril.  I am not a lawyer or legally qualified to any degree in any
> > jurisdiction, and everything I say is at least a little bit wrong.
> > Any use you make of evidence I cite, or of the ideas contained in an
> > argument I make, is without my consent or personal involvement and at
> > your own risk unless you can exhibit a signed, notarized written
> > agreement to the contrary.
> >
> > As for the expressive content of my writings, I will, sooner or later,
> > read every document that is part of the public record of any court
> > proceeding involving the GPL, and I assure you that I will notice any
> > derivative work that you create without my permission, either explicit
> > and in writing or the absolute minimum implied by my having sent
> > messages to their original recipients and implicitly to the Debian
> > e-mail archives, which are not in the public domain.  For all my
> > differences with the FSF, I will not be used as somebody's
> > stalking-horse against them.
> >
> > Cheers (to debian-legal, anyway)
> > - Michael
> Are you making a legal commitment to read a potentially infinite
> amount of public records within your lifetime?

Nah, most of them I'll read after that.  :-)

- Michael

Reply to: