Subject: Re: termless copyright and patents
From: simo <s@ssimo.org>
Date: Wed, 04 Oct 2006 08:30:30 -0400

On Wed, 2006-10-04 at 18:09 +0900, stephen@xemacs.org wrote:

> AFAICS, that is almost precisely what GPLv3 section 11 para 1 says.
> That is, you can be sued by Alice for any claim not embodied in the
> conveyed work.
> 
> The difference is that I want the burden of compiling libsection11 to
> fall on the community because it's useful and consistent with
> copyright law, while you want to simply say that any claim practiced
> in any GPL software is automatically available to all GPL software,
> which may be more consistent with patent law.
> 
> However, Bob still cannot practice others of Alice's claims without
> license.

No, Thomas position is much more subtle.
He say that he wants to treat the GPL program as it were prior art, and
any "obvious" (in patent terms) modification should be allowed. This
means that if claim 4 of patent A is not implemented in the program
distributed by Alice, but becomes obvious given claims 1,2,3 then
downstream has the permission to use claim 4 too.

I personally think that this would be too difficult to implement, and
will make things completely unclear. What about Patent B from Alice? If
it becomes obvious given claims 1,2,3 of patent A, will Patent B be
available to downstream programmers too.
How do you assess if something is obvious?
I think GPLv3 current text is better from a clarity point of view.

Simo.