Re: PROPOSAL: Evolution for GNOME 2.8

[Sander Vesik <Sander Vesik Sun COM>]

Havoc Pennington wrote:

> On Thu, 2004-06-03 at 13:13, Paolo Borelli wrote:
>  
>
> > My primary concern with this is that some people (me included, but other
> > on irc agreed) which just want to submit a couple of patches to scratch
> > their itches, often cannot be bothered to do the required paperwork even
> > if they would have no problem with the copyright assignment itself.
> >    
>
> I don't think this is a big concern; as Nat says copyright assignment is
> a good idea.
>
> What I'd worry about more is the asymmetric assignment situation for say
> OO.org (and I think but I'm not sure for Evolution), where one company
> has the exclusive right to create proprietary versions or link in
> proprietary code. Basically we're talking about a GPL loophole.
>
>  
This is implicit in any assignment, not just OOo - provided one entity 
controls the codebase, they
can freely relicence it under any number of licences they want, and the 
"external" licence being
GPL doesn't change anything. It is also very much the wrong gripe to 
have about OOo - with its
present licences anybody can make their own version into which they can 
link any amount of
proprietary code. Sun is not somehow special in this regard.

> Netscape had this loophole for Mozilla in the NPL, which gave them
> special privileges to credit them for creating the codebase. See:
> http://www.mozilla.org/MPL/NPL-1.1.html
>
> However the NPL has a time limit; the Netscape special privileges ended
> after 2 years.
>
> Of course Qt, MySQL, etc. use this as a business model. Red Hat's Cygwin
> does as well.
>
> In practice I think this is a real problem for StarOffice/OO.org and
> only a hypothetical problem for Evolution (now that Connector is open
> source).
>  
Its not - or rather, its only for people who haven't bothered to actualy 
read the licences and thus don't
know how reality looks like.

> I do support including Evolution in 2.8, however to the extent I'd worry
> about copyright assignment this is the issue I would raise.
>
> Havoc
>
>