"ISA, Dept. of Social Welfare" wrote:
> Incidentally, we looked at using URLs as identifiers, and decided that they
> were too unstable. The URL is an address. The document at that address
> cannot be identified from a URL, nor can any change to the document. A
> combination of other elements is usually sufficient to locate any document.
> The Identifier can be used for internal record management, or left blank.
In an Intranet, you can derive a location for the document by mapping
the transition from original document (in, say, WordPerfect format)
through the translation process to the final HTML or XML that is
published. In such an instance, I wouldn't bother with metadata on the
web page, but I would put a DC.Relation "hasFormat" (what is the term?)
pointing to the current version of that document on the Intranet site.
On the Internet, you have no such mechanism to track documents, and must
rely entirely upon the locator.
Alex
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|