Since Robert raised the question where we stand and what our timeline
looks like, I want to briefly recap:
* Because the attribution issue is quite divisive, I want us to
dedicate some more time to reconsidering and revising our approach.
* I'm developing a simple LimeSurvey-based survey to get a feel on
prevalent opinions regarding some of the attribution-relevant
questions from a sample of contributors.
* Our timeline will need to be pushed forward a bit, in part because
of this, but also for a number of other unrelated reasons. That said,
we want to move forward fairly aggressively given the constraints of
the re-licensing clause.
I think it's important to note that most other aspects of the proposed
re-licensing have turned out to be remarkably uncontroversial. I'm
very pleased that we've found so much common ground already. Even on
the attribution question, it seems that there is wide agreement that
for online re-use, hyperlinks to a page history or author credit page
are an appropriate mechanism for attribution. It's sensible to me, and
apparently most people, that other people's web use should be treated
very similarly to our own.
The fundamentally divisive question is whether principles of web use
can be applied to some of the other real world use scenarios we've
encountered: DVD, print, spoken versions, etc. Our established
practices don't give us a huge amount of guidance in that matter,
though many past and present GFDL-based offline uses support the case
for stronger attribution, and when this hasn't been granted (as in the
case of the SOS Children's DVD), it turned out to be controversial.
Clearly, many people feel that these media lack the immediacy of
access to authorship information that the web medium provides.
An important counterpoint is that these media are among the ones which
are the most important to reach disadvantaged communities - people
without Internet access, blind users, etc. - and that any onerous
requirements are arguably going to diminish our ability to spread free
knowledge. So, there are moral arguments on both sides. Moreover, as
I've noted, many names only really have meaning in the context of web
presence in the first place.
A compromise could acknowledge the principle that attribution should
never be unreasonably onerous explicitly (a principle which, as Geni
has pointed out, is arguably already encoded in the CC-BY-SA license's
"reasonable to the medium or means" provision), commit us to work
together to provide attribution records of manageable length using
smart algorithms as well as documenting minimally complex attribution
implementations, and permit by-URL attribution in circumstances where
we don't have a better answer yet. I worry, in this scenario, about
instruction and complexity creep over time, so the fundamental
principles of simplicity would need to be articulated well. And I want
to make sure that we don't embark on a compromise which achieves
nothing: that the vocal minority who feel very strongly about
attribution-by-name under all circumstances will continue to object,
that we will increase complexity for re-users, and that we will not
actually persuade anyone to support the approach who wouldn't
otherwise do so.
So, getting some more data on that question -- is a compromise
necessary and possible -- should IMO be the next steps, after which we
may revise our proposed attribution terms further. I hope that we'll
be able get some first survey data this week, and move quickly after
that.
--
Erik Möller
Deputy Director, Wikimedia Foundation
Support Free Knowledge:
http://wikimediafoundation.org/wiki/Donate