Despite the huge warning on the upload page whenever I check the upload log I always find several images with no info what so ever, and a lot that do have a copyright tag (usualy promo or screenshot), but no source. I know there is a image tagging project, but AFAIK they work off database dumps with means most of the images will float around for months before they pick them up (at least the ones that never make it into articles), so by the time they get the latest dump there are usualy several thousands new images to work though.
I would think it would be a better approach to catch these images shortly after whey where uploaded, that way we might even "catch" the "upload and run" users before they sign off for good, and get them to actualy fix theyr images. Maybe some of the existing RC patroling tools could be modified to detect images uploaded with no tags and source or something. There are fortunately not THAT many uploads (compared to regular edits), each page in the upload log span about an hour (using the default 50 uploads per page).
A more "permanent" solution might be something like Steve block suggested here http://en.wikipedia.org/wiki/Wikipedia_talk:WikiProject_Fair_use#Required_fields_when_uploading. That is change the upload form to include a required "source" field, and simply refuse the upload if the field is not filled in (same with copyright tag). People could naturaly just write "asdf" or whatever to bypass that, but at least then we know the lack of source is not a "good faith" oversight and send it straight to IFD (maybe even CSD, though that would require a new criterea). But that would still require people to patrol for these images, no amount of dire warnings will help if it's not enforced.
We don't have a patrol yet, but I like the idea. We could get a lot of images fixed this way.
Mgm
On 12/9/05, Sherool jamydlan@online.no wrote:
Despite the huge warning on the upload page whenever I check the upload log I always find several images with no info what so ever, and a lot that do have a copyright tag (usualy promo or screenshot), but no source. I know there is a image tagging project, but AFAIK they work off database dumps with means most of the images will float around for months before they pick them up (at least the ones that never make it into articles), so by the time they get the latest dump there are usualy several thousands new images to work though.
I would think it would be a better approach to catch these images shortly after whey where uploaded, that way we might even "catch" the "upload and run" users before they sign off for good, and get them to actualy fix theyr images. Maybe some of the existing RC patroling tools could be modified to detect images uploaded with no tags and source or something. There are fortunately not THAT many uploads (compared to regular edits), each page in the upload log span about an hour (using the default 50 uploads per page).
A more "permanent" solution might be something like Steve block suggested here http://en.wikipedia.org/wiki/Wikipedia_talk:WikiProject_Fair_use#Required_fields_when_uploading. That is change the upload form to include a required "source" field, and simply refuse the upload if the field is not filled in (same with copyright tag). People could naturaly just write "asdf" or whatever to bypass that, but at least then we know the lack of source is not a "good faith" oversight and send it straight to IFD (maybe even CSD, though that would require a new criterea). But that would still require people to patrol for these images, no amount of dire warnings will help if it's not enforced.
-- [[User:Sherool]]
WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l