Lists Home |
Date Index |
PIs were a common means in SGML systems to create links.
IDEAS/IADS worked like that originally. It was the
interpretation of some that links are not really content,
they are a process/function. It is a theoretical issue at the
very heart of darkness for hypertext systems.
From: Michael Brennan [mailto:Michael_Brennan@Allegis.com]
While tinkering with XLink and RDDL, I had an idea that I thought I'd toss
out on this list. I'd be interested to know if there have been similar
proposals in the past.
A number of specs put forward specialized mechanisms for associating
external resources with a document for a specific purpose. For example, XML
Schema gives us the "xsi:schemaLocation" attribute, XML Catalog gives us the
"oasis-xml-catalog" PI, and for associating stylesheets, we have the
"xml-stylesheet" PI. It seems like these could be unified into one
generalized mechanism by building upon XLink and RDDL constructs. You
couldn't use XLink syntax directly since these examples all must be declared
before the document element or on the document element itself. But it seems
to me that one could designate a generalized PI that mimics XLink simple
links that could achieve the desired result.
For example, the attribute "xsi:schemaLocation='http://foo/bar foobar.xsd
http://yadda/yadda yadda.xsd'" could become:
The PI <?xml-stylesheet type="text/css" href="style.css"?> could become:
The PI <?oasis-xml-catalog catalog="http://example.com/catalog.xml"?> could
This seems to me better than the current situation for the following
* Provides a consistent, extensible mechanism for these sorts of processing
* Leverages namespaces and URIs for identifying the intent and purpose of a
* Does not clutter up the document's infoset with processing directives as
Is this a reasonable idea or am I off my rocker on this? Have there been
similar such proposals?