If one document is to cross-reference another, the referenced
document should export a label that the other one can refer to. The
position of this label identifies the point in the referenced document
to which a reader should be transferred when the cross-reference is
followed. If you understand HTML, you will be familiar with
this idea, as it uses an HTML destination anchor for this
purpose. For use with HTX, this should have a form such
as:
<A NAME="xref_conclusions">any text</A>
In plain HTML, this makes "any text" a potential destination for a hypertext link, to which any other piece of HTML may refer.
What makes this anchor special to HTX is the "xref_" prefix given to the value of the NAME parameter. HTX recognises anchors with this form as potential targets for document cross-references, and this distinguishes HTX targets from other HTML destination anchors. This is important, because it gives the document's author control over which points in the document other authors may reference.
Normally, many HTML destination anchors will be used for internal navigation within a document. When the document is revised, these may be subject to change, so they should not normally be referred to by other documents. However, destination anchors with the "xref_" prefix can be chosen carefully (and named appropriately) to identify stable components of a document, to which enduring references can be made from other documents. The author is, in effect, exporting a set of labels that tell other authors which parts of his document it is safe to refer to.
We will refer to an anchor with the form above as a cross-reference target and the characters that follow "xref_" as its label ("conclusions" in the example above). As in normal HTML, HTX labels are case sensitive. Unlike plain HTML, however, they must be unique within an entire hypertext document, not just within a single HTML file.