Kelly Martin wrote:
On 9/22/05, Stan Shebs shebs@apple.com wrote:
I would appreciate it if you would take the trouble to see whether an image was uploaded in anticipation of article use. For instance, I often upload postage stamp scans in batches, and it may be a period of weeks before I write all the articles that will use the uploaded scans. You're not doing anybody a service by deleting these beforehand, nor are you incentivizing me to create the articles more quickly or to write first and upload later - more likely I'll just stop writing.
The policy Jimbo set out allowed a seven day window for a use to be found for a nonfree image. If you need an image held longer, I suggest putting a comment on the image description page so that patrollers know that is what's going on.
I've pondered on this a bit, and it occurred to me that there is an easy expedient that I think could be raised to the level of policy - namely, to have the image description page contain wikilinks to every article that the fair use image is intended to illustrate. Reviewers can then compare those to the image's "what links here" to see that they match, while as-yet-unwritten articles appear as red links.
To take an example, suppose we have an Israeli stamp depicting Ben-Gurion. Fair use might mean that we can use it to illustrate articles on Israeli philately, but not as a substitute for a free photo in Ben-Gurion's bio. So we make the description page have a link to [[Stamps of Israel]]; if the article is yet to be written, the link is red indicating that it's a future plan. If the link is blue but the page doesn't show up in "links here", or vice versa, something is wrong; the image is used in an article where it shouldn't be allowed, or perhaps the article has been edited to take out the legitimate usage - maybe a free image is now available, in case we can safely discard the fair use version, or maybe the article has been vandalized, and we want to respond by fixing the article, rather than deleting the apparently-orphaned image.
With a bit of formalized structure for this data, it might even be possible to write a script that verifies correct usage, and only warns about suspicious "fair" uses.
Stan