That makes
copyleft licences useless. Wikipedia as a whole has shown
great (some would say excessive) dilligence in avoiding copyvios. It
also maintains copious data to trace the origins of anything in an
article. This makes the likelihood of such a suit remote. It means
that there is no realistic protection whatsoever for copyleft material.
It means that any downstream user, particularly a commercial one, can
take anything from Wikipedia, and republish it under his own copyright
without any fear that it will be seriously challenged. I accept the
GFDL position to allow the material to be re-used by commercial
interests, but any commercial interest that uses it needs to acknowledge
its viral nature. Who defends that? Who defends it 20 years from now?
Baidupedia is a public website: that is, if they improve our GFDL-licensed work,
then anyone can use those derived works under the GFDL; whether or not they acknowledge it
or advertise it doesn't change that fact. Therefore, I don't see copyleft being
useless in this instance.
In theory you may be right, but unless the person who uses their
material knows that that material is covered by GFDL he needs to assume
that their copyright is valid. For copyleft to work their needs to be
an unbroken chain of attributions.
Ec