[Foundation-l] Instant Commons : INCORRECT

[Erik Moeller]

On 6/8/06, Gregory Maxwell <gmaxwell at gmail.com> wrote:

> Because our interface is so helpful, his tagging looks well formed and valid.

Maybe, but we do Special:Newimages patrol fairly well for this kind of
stuff. That said:

> Over several months various Wiki's pick it up via instant commons.

Why? Is it such a hugely popular and well-known book? How did these
wikis find out that it's on Commons in the first place, to know the
filename and use it on their own wiki?  This seems like a very
artificial scenario to me. It seems much more likely that you have
_occasional_ copyvios of _single_ files that people discover on
Commons. There will also be imports of topical categories.

Let's look at some of the stuff that is actually happening. There is,
for instance, this wiki here:
http://www.orthodoxwiki.org/Category:Images

They are copying images from Commons to their galleries. You can see
some of the images they copied here:
http://www.orthodoxwiki.org/Special:Whatlinkshere/Template:Commons

Right now, they are adding a {{commons}} template to these; they
probably have uploaded some without the template. They are also adding
licensing information, which is frequently incorrect, and they are
typically not copying stuff from the description page such as author
information, e.g., compare:

http://www.orthodoxwiki.org/Image:Jerusalem.jpg
vs.
http://commons.wikimedia.org/wiki/Image:Jerusalem_from_mt_olives.jpg

As of this writing, the above image is labeled as CC-BY-NC-SA, with no
author information. The Commons image is CC-BY and attributed to
User:Jgritz.

Correctly copying licensing and attribution information is complex
stuff. That it is copied at all above is already a pleasant surprise.
There will be many places where the _project leaders_ have no
knowledge of copyright. As Commons becomes a more widely used
resource, the number of these places will increase dramatically.

Leaving the process of sorting out license requirements to humans, as
we do now, causes many more problems than leaving it to machines
would.

> As proposed it is also a vehicle where any irresponsibility on our
> part can lead to greatly increased harm.

I simply disagree. If the file in question - a mislabeled book cover,
or whatever - is so popular and will evade copyright detection on
Commons for months, then a) our copyright processes function very
poorly, b) this fact could be exploited for any purpose regardless of
InstantCommons or not.

It seems much more likely that, if we do it right, IC will increase
our responsiveness to copyright issues, as someone discovers a copyvio
on an IC wiki, sees the backlink to Commons, and points it out to us.

Erik