Creative Commons Attribution license element

[Evan Prodromou <evan@debian.org>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I'm writing because I've just been made aware of this summary of the
Creative Commons Attribution 1.0 license:

    http://lists.debian.org/debian-legal/2004/04/msg00031.html

Let me first note that Creative Commons uses a suite of licenses, with
a number of mix-and-match license elements (Attribution, ShareAlike,
NonCommercial, NoDerivatives). So any CC license that would require
Attribution would also fall under this analysis.

Second, let me note how poorly timed the analysis is. Creative Commons
revised their suite of licenses this year (from 1.0 to 2.0), and this
list was asked to provide comment:

    http://lists.debian.org/debian-legal/2004/01/msg00241.html

Making our organization's ideas known to Creative Commons could have
meant a better suite of licenses for the 2.0 release. Instead, the
opportunity was missed. As far as I know, the above-mentioned analysis
wasn't forwarded to Creative Commons before today.

Thirdly, let me say that I disagree with most of the summary. I'll try
and cover the points in detail below:

1) Many DFSG-free licenses require that small portions of text be kept
intact or added to the source code or output of the program. The most
notable examples would be copyright notices, license notification,
warranty disclaimer, and notice of modifications made; the BSD
license(s) and the GPL come to mind immediately.

Putting conditions on modification does not make a license
non-free. We even codify this in DFSG point 4. Patch-only modification
is a condition on modification that we consider acceptable if not
ideal.

Conditions on modification are, of course, a matter of degree. If
conditions are _so_ burdensome as to make modification and
redistribution _impracticable_ -- "You may distribute modified
versions if you square the circle, jump Snake River Canyon on a
motorcycle, travel faster than light" -- then the right to modify is
for all practical purposes not granted.

But requiring attribution to the original author does not appear to me
to be that kind of burden. In particular, the license makes it clear
that you must "give the Original Author credit reasonable to the
medium or means You are utilizing". A Licensor could not abuse this
requirement by making Attribution impractical -- say, by providing a
5-terabyte name or title. Licensees are only required to give
Attribution in a reasonable way.

Let me note here that, although the Open Source Definition is not
identical to the DFSG, the OSI has approved a few licenses that have
equivalent or greater attribution requirements. Most notable is the
Attribution Assurance License template:

            http://www.opensource.org/licenses/attribution.php

This is not, of course, canonical for Debian, but it does provide some
suggestion that a group following guidelines similar to ours don't see
a serious problem with requiring attribution. Apache 2.0 also requires
attribution (in the NOTICE file).

2) I agree with this one. The intention appears to be to allow
copyright holders to avoid having their name used in advertisement, a
la the BSD, but in an opt-out rather than opt-in fashion.

However, as stated, it would indeed allow a license holder to prevent
_any_ mention of themselves in derivative works. This could severely
limit the licensee's freedom. An example would be an annotated version
of a work that critiques the writer, or an autobiography that is
revised to include critical comments or facts about the writer.

It would probably be useful to modify the license to show that the
licensor can revoke the Attribution requirements, but not prevent
other mention of their name.

3) As for the trademark clause, I agree that the trademark requirement
is burdensome. HOWEVER, considering that Creative Commons is _not_ a
party to the license, no action by that organization to overstep
trademark bounds could invalidate the license. If A grants B the
rights outlined in Attribution 1.0, no increase in trademark
restrictions by the third-party Creative Commons could revoke those
rights.

So, that's my feedback. I'd like to know what can be done to amend the
analysis and re-open this license (as well as Attribution 2.0,
ShareAlike 1.0, and Attribution-ShareAlike 1.0 and 2.0) for
consideration.

On the Creative Commons side, I'd wonder what opportunity there is to
get Debian's very tardy comments and critiques applied to new versions
of the CC licenses.

~ESP

- -- 
Evan Prodromou
evan@debian.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAxeP/ozwefHAKBVERAmHxAJ9Qna+IwzXTEXnGJm+pJD3vPdeQ7ACeNOCC
4FMc8pCK4bDxmzyefv6wlN4=
=xyLh
-----END PGP SIGNATURE-----