[Foundation-l] Fair use images

[Gregory Maxwell]

On 3/9/06, SJ <2.718281828 at gmail.com> wrote:
> I agree with Ec entirely.  Image deletion is broken precisely because
> it cannot be undone; please do not use it when deletion can be
> avoided.
>
> It is also the duty of text uploaders to describe the text's source,
> and justify its applicability to the article; nevertheless, we engage
> in discussion with editors rather than deleting insufficiently sourced
> work.  This mainly works because you can remove text from a page
> without deleting it from the edit history.
>
> Consider creating a quarantine for images that appear to be improperly
> tagged, or improperly used; removing images to that quarantine, and
> leaving them there for a reasonable length of time (a month?) before
> deleting them.  If anyone tries to remove a quarantined image, they
> must give an explanation or proper tag.

This is misleading and outright untrue in the case of older content. I
maintain a temporary archive of media I tag for deletion, some other
users do as well.. it's fairly easy to make your tagging bot go grab
the actual image.  In the case of older content, it would have made
its way into one or more image dumps (which are available for a
limited time on download.wikimedia.org, and which I and several others
maintain copies of forever).

It is true that undeletion is a little less convenient for images, but
it is untrue that it is irreversible.  It is also true that we've had
images tagged with things like "non commercial use only" and included
in featured articles for over a year.... some of these with tags
saying they would be deleted right away... and they remain completely
ignored until someone deletes them.

Generally images split into two groups, the most common case are
images no one cares about.. they are often already orphaned.. Tag
them, delete them, whatever... no one cares. Long deletion death-row
spans for these do not cause harm, but they also do no good. The other
class is images that are favored.. they are used in high visibility
places like featured articles.. There are editors who like them
because they improve these articles and some of the editors don't give
a hoot about free content. In these cases, a long death-row span can
do some good: someone might provide a source, but usually they don't..
in the more common case, a long wait is harmful.  The people working
on cleaning up our copyright status must be diligent to prevent the
warning tags from simply being removed and the article being
reinserted by people who don't care about copyright, they just want a
pretty picture.

When it comes down to it, an image was either sourced from someplace
else or created by a Wikipedian.

If really was made by a Wikipedia, but have no reason to know that
(uncommented upload) then we have the same problem as an image that
came from a commercial source (after all, the creator could later come
after one of our users.. and they user would have nothing to say
except "I'll settle").  If the user is still on Wikipedia then a week
is long enough to get them to confirm the source, if they are not then
it is unlikely that they ever will.

For images that have come from an outside source, if we obtained the
image once, we can usually obtain it again, and it is generally easier
to find a completely *new* image then to track down the
source/copyright info for an existing image which was uploaded without
any information.  When I have found source information for untagged
images, it's often been by doing a google image search on the subject
and scanning through the results, so it actually takes less time to
delete an image and replace it, then to find source information for an
existing image.


For what it's worth, in the over 20,000 media objects I've caused the
deletion of, I've only had to restore one, and I've restored 4 others
for other people.  Thats it.