On 5/15/06, Steve Bennett stevage@gmail.com wrote:
Could it be done in the ALT tag? It seems to me that there is a mismatch between those who want attribution to be readily available, and what people actually want when they click on the image. I think if people actually want to know who created the image, it's pretty easy to find out. But honestly, have you ever wondered who made an image that you clicked on somewhere on the web? Why would you care?
I do not believe an alt tag would be sufficient.
We frequently receive requests from commercial publishers for image copyright information, even with how easily it is currently provided.
On *articles* I believe our behavior is ideal in almost all cases because we do exactly what is expected: we provide an enlarged copy of the image. Most commercial news sites have the same behavior.
As far as I can tell, the problem is limited to the main page and potentially the portals. Do you agree?
So, it makes sense that the information is available to anyone who goes out of their way to look for it - but I don't see the value in forcing it down people's throats, when they're actually trying to get more information on a topic, and aren't per se interested in the image.
We can't require people to 'go out of their way to look for it'. We need to provided attribution in roughly the same manner as we do for the articles, and we need to make it reasonably accessible.
There are people who argue that our current behavior is insufficient and whom want an inline byline. But providing one runs the risk of just creating a greater mess (article authors don't get inline bylines, multiple editor images, and the requirements to provide historical data on GFDLed images). Hiding it further is just not going to fly.
Magnus' suggestion would actually be the most visible, especially if the attribution information was displayed directly on the page.
Australia (You clicked on [Australian flag.svg], created by Jim Smith) Australia is a great country... etc. The real problem is that we don't actually have attribution data in a readily packaged form.
Should read 'one of the problems with that proposal'. :) What we currently do is fine, there is no great need to have an easily machine readable copyright holder with our current framework.
On many images, there are actually several authors (due to touch ups etc). And there's no obvious way of working out mechanically who actually "created" the image.
Many? Well.. certainly not a majority. I would assume that ones with multiple editors are almost as common as ones with multiple copyright holders ... which would put us back in the same boat of having a long list of copyright holders.