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

Re: GPL, yet again. (The kernel is a lot like a shared library)

On Sun, Sep 11, 2005 at 11:20:51PM -0700, Michael K. Edwards wrote:
Michael> Are you saying these people are on record in believing that the GPL
Michael> "works" in the sense we are discussing -- forbidding the distribution,
Michael> on terms other than the GPL's, of code that uses a GPL library (or
Michael> other form of modular software component) through its published API? 
Michael> Can you provide URLs or other citations to back this up?  Especially
Michael> those that cite any actual law in support of their position?

Actually, no. I was under the impression you were saying Moglen was the only one to
consider that the GPL worked at all. I was apparently mistaken, and you were only referring
to the linking problem.

Regarding that, Lessig does support it, in either The future of Ideas or Free culture.
Rosen is skeptical about the FSF's position and seems not to endorse it. AFAIK, MacGowan
and Samuelson haven't voiced an opinion on the question. The other two (belgian law
professors) are pretty mich
convinced, mostly by my own article on the subject.

The reasoning is pretty simple. As Sean said, it is based on the understanding that the GPL
is a contract. As such, it has to be interpreted in the light of the parties intent. For
those cases where the FSF is the licensor, the FSF's intent is clear and well known. When
intent is unknown, we have to refer to common practice. That means that one has to refer to
the perception specialists in the relevant field (in this case, free software developpers)
understand things. Insofar as it is commonly held that the GPL requires that programs
linking to GPL'ed libraries be licensed under the GPL, that requirement is to be considered
part of the contract.

I shall stress that this reasoning is held in belgian law, and should be applicable in most
legal systems based on civil law. I will not voice an opinion regarding common law systems.

Michael> But no qualified source that I have yet found, other than those
Michael> directly affiliated with the FSF, seems to be willing to endorse the
Michael> "dynamic linking ban" any further than "well, the FSF makes these
Michael> claims, and it's hard to tell what a court will think."  I personally
Michael> don't think it's very hard to tell what a US court will think (at
Michael> least at the circuit court level and assuming competent lawyering) --
Michael> it's pretty clear from cases like Lotus and Lexmark that attempts to
Michael> extend the copyright monopoly to forbid interoperation are frowned
Michael> upon. 

As far as I can tell (I'm not a programmer), the "dynamic linking ban" has very few to do
with the Lexmark case for instance. Not being able to use shared libraries does not, AFAIK,
stop you from making interoperable software, at worse it hinders pure integration, not
interoperability. Let's not forget that in the Lexmark case, Lexmark was trying to stop
another manufacturer from making ink cartidges for it's printers *at all*. The dynamic
linking ban does not stop you from making programs using the libraries in the slightest. It
only puts *one* condition on it. Quite a different case. And even if you don't use the
library, you can still make a program interoperable with the target system, "just" write
your own functions and what will you. And if that is too hard/expensive, then maybe your program
would have been a derivative of the library after all...  But again, I might be wrong on

Michael> It's particularly bad news when the legal monopoly is combined
Michael> with market dominance in a given niche -- and there are a number of
Michael> sectors in which the FSF and Microsoft have a near-total duopoly, and
Michael> where neither demonstrates any qualms about leveraging its advantages
Michael> to squeeze bit players out of neighboring niches.

I'm not quite sure where you're going with your competition law references. There aren't
that any areas I can think of in which I would say the FSF has significant market power.

As to the OpenSSL case per se, it isn't clear enough for me to emit a judgement.

Yorick Cool
Chercheur au CRID
Rempart de la Vierge, 5
B-5000 Namur
Tel: + 32 (0)81 72 47 62 /+32 (0)81 51 37 75
Fax: + 32 (0)81 72 52 02

Attachment: signature.asc
Description: Digital signature

Reply to: