Development of Creative Commons licenses

[Terry Hancock <hancock AT anansispaceworks.com>]

Joachim Durchholz wrote:
> Greg London schrieb:

> And releasing under GPL is not for personal benefit?
> Go tell MySQL AB. Or Red Hat.

I think Greg's on the wrong foot here. The point isn't whether you
receive benefit, but whether the community does, and whether that
benefit is sufficient to be regarded as a significant "freedom" beyond
what is expected.

The curiosity of NC-ND, IMHO, is that it actually encodes what for many
years most people regarded as "fair use". The NC-ND terms require the
two things that most people figure are (or should be) true for ALL
copyrighted works:

1) If I don't make any money from the copies or compete commercially
with the original, they should have no complaint.

2) If I don't rip-off the work by making derivative versions that I then
try to sell, they should have no complaint.

However, minor derivations and changes are permitted by fair use (such
as quoting), and the *ideas* of a work are not copyrightable, so you
can't restrict work that is merely "inspired by" some other work.

There are hordes of websites which operate on precisely these
assumptions about copyright. They are technically wrong, but because of
lax enforcement (and the fact that most rights-holders are too smart to
make enemies of their loyal fans), they are pragmatically correct. By
and large, they do not get sued.

Because of this, it is extremely disingenuous to argue that NC-ND
"provides freedom", since it does not do so relative to the average
reader's impression of what ARR means. Indeed, doing so plays into the
hand of the Enemy: that is to say, the people who are trying to change
the public perception of ARR to be as restrictive as they want it to be,
so that they will accept legal changes to make it more restrictive (to
the extreme detriment of the public interest).

> Besides, if there's a choice between "release with some rights reserved" 
> and "do not release at all", I'd like to see the former happen. Even if 
> other programmers cannot copy and adapt the code, it can give them ideas 
> - numerous algorithms have been reimplemented using the GPL.

Now you are discussing the merits of NC/ND release models. They are the
same as for proprietary works, except that they legally permit what most
fans already expect to be true. By doing this, you are acknowledging
that the existing copyright law is too harsh, but you aren't embracing a
new model.

But Greg isn't arguing that NC/ND are *wrong*, he's arguing that they
are not "Free", and he's on target with that, IMHO.

> Finally, there's nothing wrong in releasing stuff for personal benefit 
> (as long as I'm not trying to take away more rights in the long run, but 
> I don't see that happen with freeware).

The problem here is that you want to add a source code requirement.

Why?

The only conceiveable reason is that you expect the source to be
improved. If you didn't, then why would you care if you release the
source code, and no one includes them when they pass the work on?

Otherwise, the license needn't mention it. You're free to include source
or not, as you please. The NC/ND licenses don't stop you from doing
that. They just don't stop people from dropping them if they re-distribute.

With ND they can't legally make improvements.

With NC they won't make improvements, because you'll be in a position to
charge them for their own work, or sue them for *using* their own work.

Anyone who understands those terms would rather write their own work
than contribute to yours (You *might* be able to come up with an
exception to this, but certainly not one worth the time and trouble to
compose a new license for).

>>copyleft et al are barn raising licenses.
> 
> Dunno what "barn raising" is.

"Barn raising" is an American tradition of collecting a large number of
people together (like a big party) to do large scale tasks (such as
raising the walls of beam-built barn, which are very heavy and
unmanagable for one or two people). Regrettably, large mechanized
equipment and commercial contractors to operate them have largely
displaced this heritage in the US, so the metaphor is becoming more
unfamiliar.

Figuratively, it means any collective process by which a large number of
people collaborate to produce a product too large for any one of them to
start. Wikipedia, Linux, and GNU are all "barn raising" projects because
they produced large projects, whose development cost would've been
prohibitive if developed in a proprietary model, e.g.:

Project         Estimated proprietary replacement cost (Y2K$)
---------------------------------------------------------------
Debian GNU/Linux     US$   9.05 billion
Linux                US$ 570.   million
GNU project          US$ 823.   million
---------------------------------------------------------------

(These are COCOMO estimates derived by me from data taken from a study
by Libre Software Engineering http://libresoft.urjc.es/Results )

Consider what it would take in terms of capitalization to create a
company to produce these products! For comparison (also in Y2K$, order
of magnitude estimates):

NASA Development to 1st Space Shuttle   US$  20 billion
Voyager Program                         US$   1 billion

(BTW, I lack comparable data for Wikipedia. I'd want to develop a cost
estimator based on the costs of writing large encyclopedias)

>>NC and ND are types of commercial advertising,
>>guerrilla marketing, free samples, hype generators, etc.
> 
> That's not illegitimate, I'd say.

Again with the moral judgements. There are a few people who would
disagree with you, but the point here isn't the ethics of the business
model, but the ethics of passing it off as a "community project" or as
something which "grants freedom".

> Actually there is no such thing as "free software" in the sense of "free 
> speech". Software is a thing, not a person, so it cannot have rights. 
> The FSF is going *way* beyond the normal sense of the word here.

Um. Speech isn't a person either, so clearly the use of "free" to modify
"something people have rights relating to" is perfectly sensible and
established in the language.

>>>But NC and ND are proprietary licenses. They are for the
>>>primary benefit of the creator.

Greg's off balance here, too, though. One has to make certain not
universal assumptions here to reach this conclusion. It requires that
people are good game-theoretic automatons, instead of complex, often
irrational creatures. There are a number of non-economic reasons why
people choose NC-ND licenses.

Greg rightly questions their rationality in making that choice, but it
remains true that people pick NC-ND license for reasons other than
protecting proprietary economic benefit.

Of course, the 2nd most common reason is that the author simply wants
power over other people's use of his work, and that is clearly contrary
to a principle of "freedom".

>>> And when people talk about
>>>GNU-GPL and NC-ND as if they were all part of the same
>>>thing, when people say they want to use NC because it is
>>>"half-open" then people who've been around and know the difference
>>>are going to be telling you just how wrong you are.
>>
>>I have yet to hear how wrong that would be.

About as wrong as "half frozen": technically there is a transition
between liquid and solid water, but in practice you can ignore that in
almost all meaningful cases.

In practical terms, NC licensed material does not enable a successful
sharing community to develop. It's use is more properly understood as an
extension of a proprietary "free-sample" strategy, in which something is
given away for free in the hopes of attracting more business.

Some people have observed that there are exception w.r.t. NC: there are
some religiously-motivated NC communities, for example, where commercial
is indeed regarded as "evil", so they community is actually motivated by
the NC term. Very niche, though.

So there may be some sense in which NC is merely "half frozen" (imagine
water molecules whose hydrogen bonds are vibrating vigourously), but it
still acts like ice (only when the bonds actually break can the water
flow as a liquid).

>>License break down into two simple categories.
>>The creator either licenses the work in such a way
>>that the community benefits as much as the creator.
>>OR the creator licenses the work in such a way
>>that the creator maintains an advantage over teh
>>community.
> 
> You're constructing polarity where a wide spektrum exists.

Both positions are over-simplifications, but the "spectrum" is a bigger
lie than the "polarity": the input function is continuous, but the
output function is highly discontinuous. In fact, a state change is a
very good analogy (I'm going to have to use this! :-) ).

The FSF's "four freedoms" are justified in ethical terms, but underlying
that is the fact that they represent the "necessary and sufficient"
conditions for establishing a sharing community around an intellectual
work in the absence of a coordinating authority (like a company or a
religious organization). It is, IMHO, the unconscious acknowledgement of
this reality that informs programmers' "ethical" opinions about the FSD.

It may be that the terms are not quite the same for aesthetic works, but
it's questionable, and in any case, I think NC and ND are generally the
wrong solution. Of course, we're talking about software here, so that's
a side point.

Cheers,
Terry


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