On Fri, Jan 16, 2009 at 10:07 PM, Erik Moeller <erik(a)wikimedia.org> wrote:
* For pictures, sound files, etc., there is often just
a single author. If
you are the photographer of a high resolution panorama that you've
contributed to Wikipedia, I think it's a reasonable expectation to be named
("Photo by Sam Johnston"), as opposed to being referred to as "Photo from
Wikipedia". This is equally true, I think, for articles where there is just
a single author, or for pictures which have been subsequently edited a few
times.
I would consider this an exception rather than the rule and in any case the
content author could always approach a content consumer to request
attribution. The consumer then has the option to cater to the author's
request but doesn't have to stop the presses for fear of an injunction as
giving them the option avoids any possibility for conflict. If contributors
are more interested in self-promotion than the community then they should
probably be selling on stock photo sites and writing Knols ;)
I do think the potential for internal and external conflict needs to be
carefully considered as there could be serious repurcussions in terms of
injunctions, bad will, etc.
* The attribution terms should avoid requiring
specific reference to
Wikipedia, so that it's clear that there is not necessarily a tie between
the project in which collaboration currently happens, and any future use of
the content. If someone creates a better alternative to Wikipedia where the
content is used, why should it be continued to be attributed to Wikipedia,
rather than the authors?
I was not proposing to *require* attribution to Wikipedia (indeed there
would be Wikipedians bearing pitchforks were WMF to try this on), rather
merely to *allow* it in order to foster re-use and avoid conflicts.
I think requiring attribution-by-history should be the
best practice
for heavily edited articles, at least until we more prominently point
out the author credit in the article footer.
The history for heavily edited articles is essentially opaque and claiming
that there is value to be derived from it is likely to mislead consumers.
Even if we were to provide statistics (say under a new 'Contribut[ions|ors]'
tab) we all know that edit counts are notoriously unreliable indicators and
besides, all legitimate edits are valuable.
Sam