[Foundation-l] GFDL Q&A update and question

[Erik Moeller]

2009/1/11 Anthony <wikimail at inbox.org>:
> Granted, including full change histories is overkill

Thanks for acknowledging this.

The GFDL (including prior versions) deals with author names for three
different purposes:

* author credit on the title page;
* author copyright in the copyright notices;
* author names for tracking modifications in the history section.

For its author credit provisions, the GFDL includes specific
permission to limit credit to the five principal authors. It is
evident that this permission exists precisely to avoid an eternally
inflating byline. Even then, it was clear that the covers of software
manuals did not have infinite space.

In the context of Wikipedia, authors do not receive credit alongside
the page title as suggested for software manuals. Indeed, the closest
there is in Wikipedia to the byline suggested in the GFDL is a link to
the document history close to the page title. In the context of
Wikipedia, authors are not named as part of the copyright notice.

Wikipedia contributors have therefore interpreted the GFDL in the
context of Wikipedia to allow for principal author attribution, with a
link to the history as a reasonable alternative. There is a legitimate
argument that, under a literal reading of the GFDL, any re-user _also_
has to include a full copy of the change history. The fact that author
credit and change history are combined in the context of Wikipedia is
irrelevant, as they are covered by separate provisions in the GFDL.

Wikipedia contributors, including you above, widely agree that a
requirement to include full change histories is "overkill" and would
make re-use extremely cumbersome if not completely infeasible for
heavily edited articles. No contributor has ever attempted to enforce
this provision of the GFDL, and it seems quite likely that such an
attempt would be rejected by a court on grounds of established
guidelines, practices and expectations, including site-wide copyright
terms such as http://en.wikipedia.org/wiki/Wikipedia:Copyrights ,
which contributors agree to when making edits.

The FSF agrees that this requirement is not in the spirit of the
license when dealing with heavily edited documents, or at least it
does not disagree, for it has not insisted on switching to a license
which preserves this onerous change tracking requirement. Instead, for
massive multi-author collaborations, it has given permission to use a
license, CC-BY-SA, which only requires to indicate that modifications
have been made (section 3.b), not to include a full record of such
modifications.

What we are left with, then, is to come up with attribution guidelines
in the context of CC-BY-SA which are consistent with reasonable
expectations and established practices for author credit per the GFDL.
Given its clear reference to five principal authors for author credit,
and given established practices to link to the page history (or even
the article, as per http://en.wikipedia.org/wiki/Wikipedia:Copyrights
), the proposed attribution-by-URL in certain circumstances is a fully
legally and ethically acceptable way to meet this need.

This does not close the door to improvements to attribution guidelines
and practices, including software changes, but such improvements are a
separate issue from the license update itself. That update, as I've
expressed before, is motivated by many factors, chiefly including
compatibility and - in summary - the overall complexity of the license
for re-users. The change history inclusion requirement is only one
element of that complexity.
-- 
Erik Möller
Deputy Director, Wikimedia Foundation

Support Free Knowledge: http://wikimediafoundation.org/wiki/Donate