RE: efficient XML

["Duane Nickull" <duane@xmlglobal.com>]

<Betty>Do you mean recursion or hierarchy?

Recursion:
	<Level>
		<Level>
			<Level></Level>
		</Level>
	</Level>

Hierarchy:

	<Company>
		<Name></Name>
		<Address></Address>
		<Contact>
			<Name></Name>
			<Address></Address>
		</Contact>
	</Company>

If you are talking about true recursion, I agree!  If you are
talking about hierarchy, I would not agree with a flat model.
</Betty>
Betty:

I was talking recursion.  The problem with recursive layers in parsing is
the added memory the parser must use to maintain the tree-like structure for
each extended node set.

I would favour a heirarchal approach to XML design for performance and
readability.

The methods parsers use (ie DOM vs. SAX) will also be a factor in
parsing/processing overhead.  In general, I think avoid too many levels is
the best bet.

<further_thought>
Maybe we should keep the most-often-likely-to-be-accessed nodes simple,
clean and near the root element.

Thoughts anyone - do you think this will make any impact in the overall
scheme of things?

</further_thought>

Duane Nickull