Re: EU antitrust and the GPL

[Alexander Terekhov]

Rui Miguel Silva Seabra wrote:
[...]
> > "So far, there is no evidence that open source licensors would use
> > these obligations with malicious intention trying to turn all software
> > into open source."
> 
> Other interesting quote from Mikko:
> 
>         Does the zero-royalty requirement in copyleft clause qualify as
>         restricted price fixing (or the setting of a maximum price) in
>         terms of the block exemption? The following observations speak
>         against such a conclusion:
>         -    A royalty-free requirement does not imply that the price of
>         the software must be zero. 

Then what does it imply?

>                                    Software can be priced through other
>         means than copyright royalties as well.

But other means would not price software, they would price 
something else.

>         -    Copyleft clause does not affect all further “production of
>         goods and services”. 

But the regulation doesn't say that ALL further "production of
goods and services" must be affected for a license to be in 
violation.

Software is a good on its own.

>                                It does not cover services at all and only

Nobody says that it does.

>         covers goods, which are “derivative works” of the licensed
>         software as further defined in copyright law.

That's not what the FSF/RMS says. 

http://www.xfree86.org/pipermail/forum/2004-March/004297.html

-----
Richard Stallman  rms@gnu.org
Sat, 27 Mar 2004 23:25:45 -0500

[quoting me]

    I say that XFree86 1.1 license is fully compatible with the GNU 
    GPL. A compilation is NOT a derivative work. 

[/quoting me]

You are right about the distinction.  However, the GNU GPL conditions
apply to creation of combined works, as well as to making derivatives
of the GPL-covered program.
-----

The intent to cover collective works (in addition to derivative works) 
can be witnessed in black and white from the text of the GPL itself. 
Furthermore, the GPL document, as published by the FSF, and as used 
by almost 100% of licensors, refers to the LGPL which clearly 
abuses the term "derivative work" and provides its own "definition" 
(misstated "that is to say" definition in the GPL itself aside for a 
moment).

regards,
alexander.