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

Re: GPL and linking

On 5/7/05, Batist Paklons <batist@gmail.com> wrote:
> [Note: IALNAP (I am lawyer, not a programmer), arguing solely in
> Belgian/European context, and english is not my native language.]

It's really cool to have an actual lawyer weigh in, even if TINLAIAJ.  :-)

> On 07/05/05, Michael K. Edwards <m.k.edwards@gmail.com> wrote:
> > Again, that's not how it works.  In the presence of a valid license
> > contract, one is entitled to contract-law standards of the
> > reasonableness of one's attempts to cure a breach when notified.  The
> > "automatic termination" clause is probably unenforceable in most
> > jurisdictions; I think (IANAL) few would even read it as authority to
> > terminate on inadvertent (non-material) breach, let alone on the
> > licensor's idea of breach if the licensee's (reasonable) construction
> > makes it not a breach.
> Automatic termination clauses are quite common, and generally held
> valid. It is often only what constitutes a breach that can lead to
> such termination that is disputed in court. In my opinion that is one
> of the few GPL license terms that is quite sound, only the grounds on
> which that termination happens seem extremely flakey to me.

You're quite right; I didn't really mean "unenforceable", I meant
"ineffective as a means of circumventing a court's authority to
interpret the contract and set standards of breach and remedy".  As in
the MySQL case, where the judge decided that the definitional issue
was a matter of fair dispute, and thus MySQL could not meet the
standard of "likely to prevail on the facts"; and even if MySQL's
interpretation was upheld the breach might well have been cured
(leaving the contract intact) by Progress's conduct subsequent to
notice of breach; and even if it weren't cured, MySQL could show
neither the prospect of irreparable harm nor that the balance of harms
favored it, given the conduct pledged by Progress.  Hence the already
pledged conduct would constitute sufficient remedy pending a full
trial of fact, even though the only remedy specified in the GPL is

What I really should have written is that automatic termination
clauses only affect the date from which the license is deemed to have
been terminated in the event that a court determines material breach,
but don't give the offeror or drafter any additional authority to
interpret whether a breach has occurred.  From this perspective, an
automatic termination clause isn't so much a way of strengthening the
licensor's authority to terminate as it is a declaration that the
licensee waives any waivable statutory provisions about notice of
termination in the event of breach.  It might also affect whether a
court-ordered remedy at the conclusion of a full trial includes
license termination (i. e., an injunction against continued exercise
of rights granted by the license) or merely damages for any conduct to
date that fell outside the license.

This is in contrast to "in the sole judgment of the licensor"
language, which as I understand it can only take effect upon notice in
most jurisdictions, and amounts to "termination at will" plus a
covenant not to terminate without a reasonable belief that one of the
termination conditions has been met.  Such language (which is not
present in the GPL) places the burden upon the licensee to
demonstrate, in the event of notice of termination, that the licensor
did not have a reasonable basis for belief that there was reason to

Is that how it works in your jurisdiction, more or less?

> As to the whole derivative work discussion, my opinion is that a judge
> would rather easily decide something isn't a derived work. The linux
> kernel, e.g., wouldn't need those notes of Linus to allow use of the
> API and so on, on the simple reason that the kernel is designed to do
> just that. In Europe at least one has an automatic license to do
> everything that is necessary to run a program for the purpose it is
> intended to, unless explicitly otherwise agreed to. I believe for the
> GPL to rule this out, it has to draft a clause that says: you cannot
> link to this program in such and such a way, unless it is also GPL'ed.
> In general exceptions to a rule have to be very precise, lest they
> become the rule and the rule the exception.

Woohoo.  Yes, that's how I understand it under US law as well
(IANALIAJ), with a couple of asterisks about estoppel and laches.

> I am reasoning from a legal background, and I believe that is also wat
> a judge would do. It is my general opinion, following Michael, that
> large portions of the FSF FAQ are simply wrong. I have written some
> more elaborate papers on that topic, albeit discussing intellectual
> property in more general terms, focussed on Open Source. See
> http://m9923416.kuleuven.be for that (unfortunately, the most
> interesting one is written in dutch, and I do not have time to
> translate).

I suppose that if I profess to be able to read legalese, I ought to be
able to tackle Dutch, with a little help from Google and/or Babelfish.

> Kind Regards
> Batist

- Michael

Reply to: