On 15/08/05, Walter Vermeir walter@wikipedia.be wrote:
Is there a good reason to allow direct acces by url to the images? For the Wikipedia-projects we use the [[image:foobar.png]].
Because that's how the web works: an HTML page doesn't include images, it simply refers to them; your browser then sends off requests for each image and displays them at the appropriate point on the screen. The only way of telling what page (if any) an image is being displayed as part of is to look at the "referer"[sic] header, and even that's not a 100% reliable indicator.
Or use a direct url that is only works for a short time?
An intriguing idea, but since pages are cached, they wouldn't necessarily have "up to date" image URLs in them. Also, unless you had some kind of system to randomly generate and store URL tokens on the server (which would potentially be worse resource-wise than the odd unblocked leecher) determined leechers would simply reverse-engineer the naming scheme and carry on regardless. I may be wrong, but I think this would be more trouble than it was worth.