[Foundation-l] [Commons-l] Requirements for a strong copyleft license

[Gregory Maxwell]

On Dec 2, 2007 9:42 AM, Anthony <wikimail at inbox.org> wrote:
> On Dec 2, 2007 8:57 AM, Brianna Laugher <brianna.laugher at gmail.com> wrote:
> > Is "weak copyleft" not comparable to the LGPL? LGPL appears to have a
> > place; why not "weak copyleft"?
> >
> I think the argument is specific to images, which tend not to have as
> significant of copyrightable changes made to them as software
> libraries.  Sure, maybe a newspaper cleans up an image, lowers the
> resolution, and converts it to black and white before including the
> image in the newspaper, but this is not a significant creative change,
> so the benefit of having those changes released under a free license
> is negligible.
>
> For software libraries, weak copyleft serves a purpose.  For text,
> weak copyleft serves a purpose.  For images, much less so.

*Exactly*.

Copyleft makes works somewhat *less free* but in return if it's well
applied it keeps the covered and works built from the covered work
free, thus expanding the pool of freely available works.   It's a
trade-off.

When people build new copyrightable works out of images they tend to
use them as near-verbatim components.  There are many stock photo
companies which bring in, collectively, over a billion dollars per
year licensing images for this kind of use (for example Getty Images
alone has an income of around 800 million/yr,
http://edgar.sec.gov/Archives/edgar/data/1047202/000119312507240536/d10q.htm).

Like software, most documents are written for purposes other than sale
(personal, private, special purpose limited circulation uses, etc).

This means that when we copyleft our illustrations we're creating a
substantial incentive for people to freely license works: If we had a
copylefted collection of comparable size, character, and quality to
Getty's collection we'd have a billion dollar a year incentive to
freely license works... works which wouldn't otherwise be freely
licensed because the default is non-free, or because people think
"just maybe this will make us $$$$ in the future".

Sure, the groups who make copyrighted works for the sole purpose of
selling those works for a profit aren't going to be especially
interested in such an offer. Microsoft isn't especially enamored with
the GPL.  Perhaps people will call me a zealot for not being too sad
about a failure to maximally assist groups which are locking up
content (probably forever) behind restrictive licenses. :)  Besides,
there is still a huge amount of PD and non-copyleft freely licensed
works for them to take.

Now back to weak copyleft for images.   Virtually all of that billion
dollar use is for verbatim or nearly-verbatim inclusion of images into
larger documents and articles. Usually the modifications made are so
trivial that it's easier to perform them yourself, especially since
content licenses don't have an easy way to ask for a machine-readable
corresponding source.

As a result a 'weak copyleft' wouldn't have the billion dollar/yr
incentive to freely license new works. Since it wouldn't have most of
the positive effect, I think it's a bad trade-off.   We should be
encouraging authors to use more liberal licenses (cc-by, and 'PD', for
example) if they are not interested in real copyleft, or if
copylefting that particular work will not be beneficial.