Megginson and XMLNews

Tue Apr 13 03:59:00 BST 1999

Walter Underwood writes:

 > I expect to ship our next release pre-configured for NITF,

That's wonderful.

 > but I sure would like to see some common practice beyond <title>.
 > Mostly, our customers would appreciate it, and the people doing
 > searches would get better results.

Actually, I think that you need something a little more robust --
otherwise, we'll end up with a hodge-podge of rules for what element
names people can and cannot use.  I would not want to forbid someone
from using something like this:

  <?xml version="1.0"?>

  <person>
   <title>Dr.</title>
   <firstname>Charles</firstname>
   <lastname>Goldfarb</lastname>
   <desc>Originator of SGML.</desc>
  </person>

Universal names (as in "Namespaces in XML") get you part way there,
because different document types can share semantics of well-known
element types:

  <?xml version="1.0"?>

  <book xmlns:html="http://www.w3.org/Profiles/xhtml1-transitional">
   <front>
    <html:title>This is the book title</html:title>
   </front>
   <body>
    [...]
   </body>
  </book>

What's really useful, though, is to develop some kind of inheritance
scheme, so that you can say "this is just like an html:title, except
that it's also a little more specialised".  Architectural forms
provide a very lightweight mechanism for this; XML Schemas will
probably provide another.

Personally, I'd love to see NITF take advantage of namespaces, even to 
a very small extent.  To start, a simple default namespace would be
nice:

  <?xml version="1.0"?>

  <nitf xmlns="http://www.iptc.org/iptc/namespaces/nitf/">
   <head>
    <title>Simple Story</title>
   </head>
   <body>
    <body.head>
     <hedline>
      <hl1>Simple Story</hl1>
     </hedline>
     <byline>
      <bytag>By David Megginson</bytag>
     </byline>
    </body.head>
    <body.content>
     <p>This is a simple story that mentions <cite>Shakespeare in
      Love</cite>.</p>
    </body.content>
   </body>
  </nitf>

This would allow other document types to reuse NITF components in a
well-defined way, and search engines to recognise them wherever
they're used.  Right now, we're not doing this in XMLNews-Story
because we want to remain strictly subset-compatible with NITF, but
we'll certainly encourage the NITF people to consider updating the
spec.

In fact, since NITF borrows heavily from HTML (and also a bit from
HyTime, though that part is not included in the XMLNews-Story subset),
it would be nice to put the HTML stuff in a separate namespaces so
that search engines and other processing software can do something
useful with it even if they do not know NITF itself:

  <?xml version="1.0"?>

  <nitf xmlns="http://www.iptc.org/iptc/namespaces/nitf/"
	xmlns:html="http://www.w3.org/Profiles/xhtml1-transitional">
   <html:head>
    <html:title>Simple Story</html:title>
   </html:head>
   <html:body>
    <body.head>
     <hedline>
      <hl1>Simple Story</hl1>
     </hedline>
     <byline>
      <bytag>By David Megginson</bytag>
     </byline>
    </body.head>
    <body.content>
     <html:p>This is a simple story that mentions <html:cite>Shakespeare in
      Love</html:cite>.</html:p>
    </body.content>
   </html:body>
  </nitf>

This might help a bit with the search engine problem.

All the best,

David

-- 
David Megginson                 david at megginson.com
           http://www.megginson.com/

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)