[Fedora-legal-list] Linking an LGPL library statically to an GPL program

[Joerg Schilling]

Orcan Ogetbil <oget.fedora at gmail.com> wrote:

> On Tue, Nov 17, 2009 at 8:29 AM, Joerg Schilling wrote:
> > Orcan Ogetbil wrote:
> >
> >> Yes but you are missing one thing. The library is LGPLv2. It is not LGPLv2+.
> >> Doesn't it make the resultant binary GPLv2, without the + ?
> >
> > There is nothing in the GPL that requires you to put binaries under GPL.
> > In fact, you can't even do this in many cases. You just need to follow the
> > conditions in section 3 for the binary.
> >
>
> That assumes that the binaries are not considered "derived work", on
> which there is no general consensus. There are opinions in both ways.

I would be interested in getting information about lawyers who claim that the
binary compilation results may be a "derived work" and of course I would like 
to read their legal reviews on this topic in order to be able to compare 
claims. I did not yet get in contact with such people....

If you talk to various lawyers from vaious countries, you will see that there
seems to be a consensus that in order to create a derived work, you need to add
a suffifient amount of own creation which only happens if you modify sources. 
It is obvious that this does not happen with the automated process of compiling 
and linking.

While I was trying to sue companies who abuse GPLd software in 2001 (long 
before Harald Welte did the same), my lawyer explained me that few of the claims
of the GPL are enforcable in court and independend lawyers from different 
countries confirm that this does not only apply in Germany.

The US lawyer who counsels the OpenSource Initiative explains that the 
combination of a work under GPL and another work that is a library is a 
"collective work" in contrast to a "derived work". See page 114 ff. in:
http://www.rosenlaw.com/Rosen_Ch06.pdf

BTW: I asked Eben Moglen about this in 2001 and in 2008 and he confirmed that 
the automated process of compiling/linking does not create a "derived work".


> > If you ever like to convert LGPLv2 code to GPL, you of course cannot convert it
> > to GPLv2+ but only to GPLv2.
> >
>
> Now this contradicts spot's conclusion. Where is the catch?

It is obvious that your original conclusions are correct. Why should
an author who forbids to use a later version of the LGPL (the license he 
originally selected for his work) give permission to use a later version of the
alternate license (GPL)?

BTW: I cannot speak for the legal situation in the USA....
But in Germany, the law forbids you to sign a contract that you don't know 
while you sign. If the last publish date from the Copyright holder was before
the final version of the GPLv3 was finished, the phrase "GPLv2 or any later"
is void and GPLv2 applies. 


Jörg

-- 
 EMail:joerg at schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
       js at cs.tu-berlin.de                (uni)  
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily