Subject: Re: For Approval: CeCILL (providing source)
From: Stephane Dalmas <Stephane.Dalmas@sophia.inria.fr>
Date: Fri, 16 Sep 2005 15:27:35 +0200

Wilson, Andrew wrote:

>Stephane Dalmas wrote:
>
>  
>
>>>Andrew Wilson wrote:
>>>Having freshly
>>>re-read the license, now I'll add (3) the excessively broad patent
>>>covenant not to sue, which is not limited to the actual
>>>      
>>>
>CeCILL-licensed
>  
>
>>>work or its derivatives.
>>>
>>>      
>>>
>>See my previous answer on the first two points. The patent clause seems
>>    
>>
>>perfectly fine for such a license (in the spirit of the GNU GPL).
>>    
>>
>If there is an explicit patent grant in GPL 2.0, it's certainly
>news to me.  
>
I wrote "in the spirit of the GNU GPL" which is quite different from "as 
in the GNU GPL".  I guess you know that the FSF is not fond of software 
patent and would (I believe) not regard this kind of clause as a problem 
at all.

>Your patent grant in CeCILL, however, is so broad that
>as written it raises the possibility of "patent laundering," i.e.,
>Licensor's patents which are completely unrelated to the covered code
>may be inadvertently licensed by releasing code under CeCILL.
>  
>
This simple idea that some valuable patent may be "inadvertently" 
included in a code exemplifies why people may think that software 
patents are not be a good idea at all.... I can't really imagine a 
useful invention that has been hard to discover and that can be 
inadvertently included in a source code by a programmer. But that is of 
course another debate.

>Please see the patent grants in licenses such as EPL, CDDL, and
>Apache 2.0 for more broadly acceptable terms which do not allow
>"patent laundering."
>
>  
>
We made our homeworks and are well aware of the various patent grants in 
several licenses. We deliberately chose to word it as it is (and not as 
a patent license because some formalism and special obligations may be 
attached to patent licenses, this is the case in France, for example).

>>Internationalization of licenses is a very difficult problem but it is 
>>not a reason not to try to solve it. CeCILL is a first attempt for 
>>Europe and we plan to go ahead.
>>    
>>
>
>Internationalization is good. Vanity licenses are bad.  If you are
>arguing that GPL is not enforceable in the EU and a new license
>is required, there are multi-million lines of European-written GPL
>code which would seem to prove that the free/open source software
>community of the EU does not agree with you
>  
>
If you are arguing that the fact that all these programmers releasing 
code under the GPL make the GPL enforcable then you are wrong. What 
enforces the GPL are judges and national laws. Fortunately or not, 
programmers have little power on that no matter how much lines they 
write. And besides you are not asking the right question. The question 
is not really whether the GPL is enforcable or not (althougth it may be 
a real question in some jurisdictions) but what exactly is enforcable in 
the GPL and to what extend. As you may know there was a case in Germany 
where some part of the GPL were indeed enforced but the judges said that 
they were not sure at all about some other part of the GPL (with regard 
to German law). We never pretended we wrote CeCILL because the GPL was 
not enforcable in Europe but because we wanted to have a GPL-like 
license that will offer more legal safety by being "more enforcable" 
than the GPL in Europe at least.

    Stéphane.