vs CC BY-SA 4.0

[Richard Fontana]

On Thu, Jul 3, 2025 at 4:17 PM Bradley M. Kuhn <bkuhn at ebb.org> wrote:
>
> Winston de Greef wrote:
> > I was wondering why I might consider this license instead of something like
> > CC Attribution-ShareAlike 4.0 International.
>
> The key reason that I recommend avoiding CC-BY-SA-4.0 is that it is *not* a
> true copyleft.  Specifically, CC-BY-SA-4.0 (and other CC “share alike”s makes
> basically no effort to assure that downstream recipients can *effectively*
> take advantage of their right to modify.
>
> Traditionally, software-focused copylefts always included some specific
> description of what materials and rights must be provided downstream to allow
> users/consumers to make *effective* use of their rights.  For example, GPLv2
> talks about “complete, corresponding source code” which includes “scripts
> used to control compilation and installation of the executable”.  A different
> copyleft license, GPLv3, also has a lengthy definition for its defined term
> “Corresponding Source”.
>
> IMO, this material is the heart of what it means to be copyleft: if you can't
> use the license to demand the artifacts you need *easily* engage in your
> right to modify and/or install the software, then you don't really have a
> copyleft license.
>
> CC-BY-SA (any version) has never any real equivalent of a “source code
> provision”.

I would also argue that CC-BY-SA (any version I'm aware of) is not
truly (in general) a _libre_ license, despite common assumptions
otherwise. One particular problem that exists in all Creative Commons
licenses I'm familiar with is the language about patents - in CC-BY-SA
4.0:
"Patent and trademark rights are not licensed under this Public License."
While in general what a FLOSS license should say, if anything, about
patents is a difficult topic, I would contend that a license that says
"patents aren't licensed" can't be considered _libre_ (trademarks
don't raise the same concern).

Admittedly, for many kinds of works commonly licensed under CC-BY-SA
4.0 and other Creative Commons licenses, this is not likely to matter
because such works will, probably, not be susceptible to being covered
by a valid patent claim. Occasionally, however, purportedly FLOSS
projects use CC-BY-SA and other CC licenses for code and other
functional material, possibly assuming that they are _libre_. Similar
language is also in CC0, which *is* pretty commonly used for code.

Still, we can certainly learn from studying how Creative Commons
attempted to solve copyleft and other libre or quasi-libre license
implementation problems.

Richard