Probably the absense of a robots.txt file at download.wikimedia.org is a expected behaviour, but it is resulting in bandwidth waste: Google is indexing the .gz files, Yahoo the .xml ones
http://www.google.com/search?q=some+site%3Adownload.wikimedia.org
http://search.yahoo.com/search?p=some+site%3Adownload.wikimedia.org
Version: unspecified
Severity: enhancement
URL: http://download.wikimedia.org/robots.txt