during my work on a kind of type dictionary in xml format
some problems occured, that cannot be solved for know.
you specified an element id attribute or some global key in a xml schema defintion.
you have several documents (x,y,z) who include the same document (a) of the same type.
till now all things work and no duplicated ids / keys occure.
but what if we have another document (b) which includes all of them.
in the best case we get 4 duplicate ids, which means that document
was included multiple times.
the same problem is known in context of programming languages and there it is solved
in terms of defines or managed sources. it would be nice if an additional attribute of "include" would
prevent such multiple recursive includes.