I understand Milos' concern, actually, and this is the most reasonable
objection to a URL link for attribution purposes that has been raised so
far. It is true that the Internet is by its nature impermanent, evolving
both in content and in structure. It's by no means guaranteed that if we
in a printed book in 2009 it will still be accessible in 2019.
On the other hand, if we printed out the names in the book... then as long
as you have the book you have the names, because they travel together. We
may change the syntax of the history link, the most common method for
locating content on the web may change (either structurally, or because of
device evolution), or the sites might for some reason come down. We should
also consider that ideally we want our content to be usefully credited in
areas of the world where Internet access is very limited, or where Wikipedia
is specifically blocked. Thinking ahead, these are the parts of the world
most likely to be using a paper Wikipedia anyway.
I do understand that there are mediums where this is impossible, and I think
perhaps the solution requires an outline that describes different (but
reasonable) standards based on medium category, broadly interpreted.