A simple solution would be to create new categories, such as Geograph:West Sussex and Geograph:Barnes.
Bots could then upload images to those, which could be subcategories of the respective parent categories, without clogging the latter, and images could be switched manually, as they're checked (simply by deleting the prefix were applicable).
On 14 April 2012 14:30, John Byrne john@bodkinprints.co.uk wrote:
Yes, many of us are aware of the issues with Geograph, above all WSC.
I agree the categorization side of it has been the real Achilles heel, and in my experience the problem is often worse than WSC suggests. When I filled up the Commons category for Wimbledon Common, http://commons.wikimedia.org/wiki/Category:Wimbledon_Common, I found that a significant number of images were categorized in "West Sussex" categories (what, 80 miles away?) and several others as "Barnes" (only 5 miles out, but that's a lot in London). But the good news was that I was able to find these images easily enough through the basic Commons search, as the original Geograph text info had enough detail. I've had this sort of result doing other categories.
I understand that because templates were mostly used to record images as uncategorized etc, and categorizing with cat-a-lot doesn't remove these, and they are a pain to remove when you're doing bulk, these tend not to get removed. So a good number of the images categorized with uncategorized or category query templates are actually ok, and we don't have any reliable numbers for what is still a problem. Many of the ones supposed to have problems don't, and many of the ones supposed to be ok aren't.
If you want images for a place in the UK, you should always do a basic search as well as looking at the category. But actually that's true of most things on Commons.
Johnbod
Wikimedia UK mailing list wikimediauk-l@wikimedia.org http://mail.wikimedia.org/mailman/listinfo/wikimediauk-l WMUK: http://uk.wikimedia.org