For Approval: GPLv3

[Chris Travers]

Hi all;

Mahesh T. Pai wrote:
> Alexander Terekhov said on Wed, Aug 29, 2007 at 12:19:01PM +0200,:
>
>  > You're still missing the point: the GPL states that your recipients
>  > get the license from the
>  > original licensor (and that means BSD, not GPL), not you.
>  > 
>  > Is there GPLv4 in the making to fix that bug? Just curious. The fellow
>  > from softwarefreedom.org posting here might want to elaborate...
>
> If the point you are trying to make is that BSD and GPL are not
> compatible because people distributing others' BSD covered works
> cannot apply the GPL to such distribution because licenses under GPL
> get the license from the original copyright owner, while the
> receipient of BSD covered work get it from the sublicensee, well, you
> are correct as to the conclusion, but in error as to consequence.
>
> In practise, this sublicense distinction makes no difference when it
> comes to remedies or enforcement.
>
> I am open to correction on this. 
>   

It might make a difference by my reading of Gates v. Bando and the cited 
standard for derivative works on which the court derived additional 
tests.  The basic argument is that the author must show copyright 
ownership for the portions copied into a derivative work.

IANAL, of course.

Hypothetical:

Company A releases a product under the GPL, but includes three files 
under the BSDL.  Although the files are unaltered and the BSDL notice 
remains the files are listed as licensed under the GPL.

Company B takes those BSDL files, assumes they are unaltered, and uses 
them in another propietary product.  One of Company B's developers 
states that the files were copied from the GPL distribution, and company 
A sues arguing that the proprietary product is an illegal derivative 
wrok.  Company B's defense is that they have not infringed on any of 
Company A's copyrights and therefore company A both lacks a claim of 
infringement and lacks standing to bring a claim.  They then only after 
the fact go back and confirm that no changes have been made to these 
files.  Since the GPL states that acceptance is not required of the 
contract, it would seem to be a useful defense.

I am open to contradiction here.

Hypothetical:
Same case as above but in the subsequent check, substantive alterations 
are found (let us say, two functions were rewritten almost entirely).  
In my view, this would bring the GPL into play because company A 
licensed those *changes* under the GPL and company B misappropriated 
them without permission.

In short "relicensing" may not extend the GPL's reach into original BSDL 
fragments without alteration.  The main difference is that the 
conditions of use under the BSDL don't preclude its use the BSD code 
(perhaps even with minor alterations) under the original BSDL.  In 
essence one can only add additional restrictions to BSDL code by adding 
copyrighted components (including but not limited to code) which are 
encumbered by further restrictions.

The question about the GPL3 is "what does it mean to remove additional 
permissions?"  Does this mean "allow for encumbering changes to be 
added?"  Or does it mean actually to remove permissions granted by other 
licenses?  In the first case, the GPL3 may be compatible with any 
license that grants a superset of rights and does not extend to 
derivative or collected works (that would seem to include the MS-PL), 
and in the second case, it is incompatible with every other license out 
there.

Best Wishes,
Chris Travers


-------------- next part --------------
A non-text attachment was scrubbed...
Name: chris.vcf
Type: text/x-vcard
Size: 171 bytes
Desc: not available
URL: <http://lists.opensource.org/pipermail/license-discuss_lists.opensource.org/attachments/20070829/2c04d2aa/attachment.vcf>