Re: Distribution of media content together with GPLv2 code in one package?

[Rudolf Polzer <divVerent@alientrap.org>]

On Mon, Apr 05, 2010 at 09:49:36AM +0100, Matthew Johnson wrote:
> I think it's pretty clear that you can distribute data alongside a GPL work
> without it being covered by the GPL. The reason why program objects which link
> the GPL work need to be distributed under the GPL is that the result
> constitutes a derivative work. If it did not, then copyright law cannot cover
> said aggregation. Whether linking constitutes a derivative work is something
> which has not been tested in court, but since that is the FSF's position, that's
> what most people go with. There's never been any suggestion that data which is 
> processed by the application is covered: otherwise you couldn't choose the licence
> of documents you created with a GPL word processor. More importantly, however,
> the GPL contains a clause permitting 'mere aggregation'.

The question rather is:

what exactly is mere aggregation?

Is "the game" not ONE new product, and thus one single item? After all, users
don't download an engine and data for it, but they download "a game".

Also, the data typically is packed into "somewhat opaque" file containers. In
the specific case, these are renamed zip files. To perform a separation, an
idea would be using multiple of these files, one for files of each license. But
would that be separated well enough?

Also... as "the game" will not run without that data - can it even be called
mere aggregation?

Best regards,

Rudolf