Following on Ben Hammersley's examples, I created an XML namespace illustration, which follows.

The first block has no namespaces. The second block uses namespaces to clarify that the elements are a catalog item on a book about Duke Ellington. The third block uses namespaces to clarify that the same elements are in fact an item for a meeting with British royalty.

<item>
  <title>The Duke of Cats</title>
  <subjects>
    <term>Duke</term>
    <term>cats</term>
  </subjects>
</item>

***

<catalog:item>
  <book:title>The Duke of Cats</book:title>
  <book:subjects>
    <jazz:term>Duke</jazz:term>
    <jazz:term>cats</jazz:term>
  </book:subjects>
</catalog:item>

***

<meeting:item>
  <royal:title>The Duke of Cats</royal:title>
  <royal:subjects>
    <British:term>Duke</British:term>
    <animal:term>cats</animal:term>
  </royal:subjects>
</meeting:item>

I created this illustration for myself as a way to think about the relationships between XML document schemas and namespaces that might be described by their own schemas. But, I also thought that this illustration might be useful in general as one way to demonstrate what namespaces can do to "meaning" in XML documents.

(I didn't feel up to doing it myself, but I would be interested in seeing these examples expressed in RDF/XML.)

In the iCite net, the relationship between elements and schemas and namespaces is all basically within the concept of schemas. With XML, schemas and namespaces are optional, and, I think, this reflects the XML design orientation towards data fragments—for example, as appropriate for applications like data interchange.

Anyway, this is namespace stuff I have been thinking about.