From: Saa�ha Mets�rantala
Thanks for the comments, Saa�ha.
> Having numbered levels may result in some difficulties to insert any
> extra step between two levels,
I don't understand what you mean by inserting steps between levels.
> to get the INTERSECTION of the elements at the same level, which I
> consider is not consistent with the preceding example.
That's right, and as I pointed out, there are loose ends to tie up and semantics to tighten, but before doing that we should first nail down the use case. For example, an acceptable rule might possible turn out to be that if regions at the same level intersect, then we are looking at the intersection, and if they don't, we are looking at the union. On the other hand there may be a use case where two regions do intersect but we want the union, in which case that rule would not be acceptable. But we do want to keep this as simple as possible and not introduce complexity that isn't supported by use cases.
> one question here is whether we should rely on the order of the
> elements within hierarchicalGeographic or if we should rely on their
> nesting (or both).
I don't understand what you mean by nesting (as there hasn't been any nesting introduced in any of the examples).
As to relying on the order of the elements ... I think the MODS EC is reluctant to rely on preservation of element order, in cases where the XML standard doesn't require it. We understand the argument that there is no known XML software that does not preserve element order, but that doesn't mean that all future XML software will preserve order.
> One suggestion would be ....
> ... adding the geographicUnion element when needed.
The EC has talked about a number of possible approaches, including the introduction of a <union> and <intersection> element. However, as I said, I think we need to see concrete use cases to justify the significant complexity this would introduce.