Jimmy Wales wrote:
I couldn't possibly disagree more, I think. *Disallowing* anon edits would not be in the spirit of the GNU FDL. I just don't know where this idea is coming from that the attribution requirement extends beyond the identification that someone actually gives us. It certainly doesn't come from the law, or from the license itself.
I'd tend to agree with this interpretation --- my reading of the GFDL is that it's informally saying "you have to keep the original attribution information, with the exception that you don't have to keep more than five authors". If somebody published a book anonymously under the GFDL, then presumably you could make derived versions of that book with no author attribution. At the very most, you could say "originally published anonymously." Which I think is the same situation here -- "this article contributed to by User2, JonDoe, and several anonymous authors."
-Mark