Development of Creative Commons licenses

[Terry Hancock <hancock AT anansispaceworks.com>]

Gregory Maxwell wrote:
> On 2/12/07, Terry Hancock <hancock AT anansispaceworks.com> wrote:
> They would likely be concerned if the different but still free license
> was a weaker copyleft than the license they used. There is no point in
> having a strong license if by some license slight of hand a downstream
> user can switch to a weaker license, since downstream folks who want
> to circumvent the stronger terms will do exactly that if the option
> exists.

ISTM you're missing a subtley in this particular thread. The idea is
that there is a third state, between "mere aggregation" and
"derivation", which involves "semantically linked" works.

I still don't know if that can be written in a legally-binding way, but
it would be useful.

In the case of music for a movie soundtrack, the music's license *would
not change* just because it is synched with a movie under a different
license. But the movie wouldn't otherwise be required to use the same
license.

The closest extent analogy would be to the LGPL which allows a specific
kind of linking (dynamic or static linking of a software library to a
package). The Library remains under the LGPL, and any changes to it must
be treated in the same way as the GPL, but this specific type of
distribution may be used with proprietary works.

In the case we are envisioning, the requirement is slightly different:
the linking may occur outside of the object's own license, but it must
still be under an "approved" free license.

IOW, you can imagine a variant of the LGPL which allows linking of
libraries to *ANY* free software license (say, "Any OSI Approved
License"), but NOT to proprietary licenses. This would be more
restrictive than the existing LGPL, but still not as restrictive as the GPL.

This would allow, for example, a GPL-like library to be linked to MPL
code (which is otherwise incompatible), but not to be appropriated by
Microsoft into a commercial/proprietary product.

Note that GPLv3 makes it easy to declare such a license using the same
Section 7 provisions that are used for the new LGPL.

> This is the conceptual flaw in the compatibility approach currently
> taken in the cc-by-sa-3.0 draft:  any work protected by a license in
> the pool of licenses compatible with cc-by-sa-3.0 receives protection
> only as strong as the weakest license in the pool.

Yes. That's a problem. I think certain steps can be taken to mitigate
this risk, though. One is this idea of keeping separate components under
their separate licenses, but allowing them to be linked together.

> This might be acceptable if the most interesting failure mode of a
> free content license was to be too restrictive. But because license
> restrictions can be fixed after the fact (by issuing new licenses)
> while excessive permission can not be fixed except for new
> works/derivatives, the too restrictive case is not the one we should
> be solving.

Yes, the insufficiently restricted case is more of a hazard.

I think automatic upgrade clauses are the real kicker. If a license
containing such a clause is declared "compatible", then the way is
opened for a bad license steward to essentially leak the whole commons
out of copyleft entirely. Which would be *BAD*.

Cheers,
Terry

-- 
Terry Hancock (hancock AT AnansiSpaceworks.com)
Anansi Spaceworks http://www.AnansiSpaceworks.com