[cc-community] What does NC means?
Terry Hancock
hancock at anansispaceworks.com
Fri Sep 21 10:44:02 EDT 2007
Mike Linksvayer wrote:
> I have some anecdotes -- sorry, the only two I can think of off the top
> of my head are CC family, so don't really count:
> http://joi.ito.com/archives/2004/05/27/creative_commons_20.html
> http://joi.ito.com/archives/2004/06/28/why_free_for_commercial_use.html
> http://laboratorium.net/archive/2007/02/16/share_and_enjoy_25
Though, of course, a few anecdotes does not prove a trend!
(Thanks for the links, though).
> And three time series data points, which don't count for much:
> http://creativecommons.org/weblog/entry/5936
I've already worked some with these numbers. The breakdown by "license
module" is practically useless information, though, because the license
modules are NOT orthogonal.
E.g. the success or failure of "NC-SA" has nothing to do with the
success or failure of free licensing, but "SA" does. OTOH, "By" (alone)
does affect "free", but almost all uses of "By" are in combination with
something that doesn't. So you're grouping a bunch of non-correlated
licenses together because they share modules, and breaking up licenses
that do have commonalities.
However, a breakdown into "Free" and "Non-Free" licenses, with the
"Free" group including Public Domain, SA, By, and By-SA for versions 1-3
of the license, and "Non-Free" including the others, does show a slight
improvement in "Free/Non-Free" over the interval considered (from 2004
to mid-2006), based on CC's own link analysis figures.
Specifically, I fit power-laws to the Free and Non-Free datapoints, and
I found that a powerlaw of about "0.61%/day" fits the free licenses,
while one of "0.53%/day" fits the non-free licenses. Also, "copyleft"
(that is, SA or By-SA) is winning slightly against non-copyleft
(0.62%/day versus 0.60%/day).
At all points, of course, the non-free licenses are a considerably
larger population than the free licenses, but the free licenses are
actually gaining slightly w.r.t. the non-free licenses.
Even if this trend is accurate (caveats below), I don't know whether to
call it "success". The growth of free vs non-free is extremely subtle --
it's hardly grounds for predicting that free licenses will supplant
non-free.
OTOH, it would refute claims that have been made by free licensing
advocates that non-free CC licenses are "stealing" potential license
users away from free licenses. The data much more strongly suggests that
the CC licensing population is being primarily drawn from the general
population of licensors (and thus primarily from users of even more
restrictive non-free licenses, mainly "All Rights Reserved").
The fact that "free" CC license use is growing exponentially is good
news that ought to be considered without reference to the non-free CC
licenses (it's not like we didn't already know that most publishers and
authors today want to use non-free licenses!).
Some caveats:
1) With the numbers
- They come from CC, which is an interested party
- They are based on links to the licenses, but these
could be diluted by descriptive references rather
than license indication
- They are based on search engine statistics, and
CC switched (from Yahoo to Google, IIRC?) during
the time series, this introduces a systematic jump
somewhere in the middle of this series. Increasing
numbers could easily just mean better indexing, not
increasing numbers of links
2) With the curves
- The number of datapoints is very small, so there's
no way to factor out random variations accurately
- The assumption of exponential growth is primarily
theoretical there's not enough data to distinguish
between different to prove that the growth is
genuinely exponential. However, the growth is
definitely super-linear and it's easy to fit
the existing data with an exponential powerlaw.
3) Assumptions
- This does NOT prove the hypothesis that CC non-free
license users typically or even often switch to CC
free licenses. It is just as consistent with the idea
that both populations are growing directly from the
general population
Theoretically, we'd expect to see three possible curves fit this kind of
data:
1) Exponential -- means that growth is unrestrained. As more people
use the license, it becomes better known, and so even more use it. It's
a "snowball effect".
2) Asymptotically approaching a limit -- there is something severely
restricting growth. Either you're running out of would-be users, works,
or total web pages (for example, this would be bound to happen when all
websites in the world link to CC licenses, or when every possible CC
license user is already using it, etc)
3) Linear -- growth is occuring, but is limited by outside forces which
have reached equilibrium with your promotional forces.
Taken together, these form what's called a "sigmoid curve" which applies
to population growth in general. Initially growth occurs without limit
(resources are essentially infinite), then environmental factors
gradually slow down growth to linear. Finally, resources become
exhausted, and the growth caps out.
That CC licenses are growing exponentially is about the best you could
hope for -- it means that we are a long way from saturating the
population of licensors open to using them.
BTW, 0.5%/day is pretty fast. It's more than 600%/year!
BTW, my research indicates generally that free culture (including
free software, open source, etc) is growing exponentially, with no
sign of tapping out. So a lot of the fears about industry blocking
free culture, etc, is unfounded. If "the industry" is trying to
block free culture, it's failing pretty badly -- there's no evidence
of an effect.
Just the slow (or not so slow), gradual build of numbers. It's not as
dramatic as following epic battles won and lost, but it's what actually
counts in the end. More and more people are getting sold on the idea.
And the doubling time is only a few months. At this rate, the landscape
will be completely different by 2020 (sometime around then, IIRC,
Wikipedia will consume all available disk space on Earth ;-) And yes, it
is presently outstripping Moore's Law growth of storage media!).
Seriously, it will *have* to start tapping out before that, but
you get the picture. Similar numbers apply to desktop Linux use, etc.
Cheers,
Terry
--
Terry Hancock (hancock at AnansiSpaceworks.com)
Anansi Spaceworks http://www.AnansiSpaceworks.com
More information about the cc-community
mailing list