• From: Len Bullard <len.bullard@u...>
• To: Laurens van den Oever <laurens@x...>, xml-dev@l...
• Date: Fri, 15 Feb 2008 12:43:48 -0600

Hello Laurens:

It seems you are doing what we did with SGML DTDs that applied 'content
tagging'.  SGML required the DTD so it was used to parameterize the editor.
The question was then a matter of linking stored fragments to create
composite documents in accordance with the master type definition given
there was no namespace specification.

The IETM community was particularly advanced in the publication of aggregate
documents using relational database sources and DTD-aware editors.   I wrote
some of the DTDs and have been a semantic content advocate since that time.
It isn't that difficult and is still common practice in server side web
applications.  

My question is more directed at why an XML document database would be more
effective given the relational db can do the job.   There was quite a bit of
debate about that in the early 1990s.  Semantic structures weren't hard to
combine.  Nested numbering structures (eg, 1.1, 1.1.1, etc.) were a bit
harder specifically because they weren't topical or semantic.  HTML was a
huge letdown at the time because of the crudeness of the structures.  Links
were to strict position anchors with no semantic validity and so forth.

So far, we can achieve what is necessary with relational systems and
semantic-aware editors.  That means it isn't much more than what forms do
given the form submit process has the database semantics in the form of the
SQL statements.

XOPUS, AFAICT and I haven't looked at it, seems to be another turn of the
wheel of the content tagging systems where legal or valid combinations
depend on one or more schema.

len


From: Laurens van den Oever [mailto:laurens@x...] 
 
> Len Bullard wrote:
>
> I'd like to hear the 'compelling use cases' where enhanced semantic 
> support is a fundamental predicate of the case.

Since this is our core business, I can think of many such cases, but
I'll stick to two:

1. A consumer electronics manufacturer uses our software to create their
annual reports. These reports need to be published as soon as the
accountant approves their numbers. So they have two teams working on
this simultaneously:
- The financial division, they work to validate the numbers and get 
  approval.
- And a team of authors, they write the text based on preliminary
numbers 
  including all tables and graphs.

The challenge is to combine all this information automatically. They no
longer wanted to spend weeks to copy/paste the numbers once they are
approved.

This is achieved by having the authors use Xopus to create a structured
document that points to the relevant fields in their financial system.
This allows the authors to have full control over the document,
including adding columns to tables with explanations, reordering data
for clarity, etc. The controlled environment maintains the links with
the financial system. So when the numbers are approved, they can be
integrated in the document with one mouse click. The only manual work
left is going over any last minute changes in the numbers that should be
reflected in the text.

The implementation of this process gave their report a boost in quality
while both production cost and time to market went down.

2. An education publisher uses our software to create their content.
This content drives their e-learning environment but is also used for
their printed material. Their content is based QTI while the output is
based on SCORM. On top of these requirements they are working with many
external authors who were previously working with a word processor. 

So the requirements imply complex mixed content document that has to be
created by non-technical people who can not be trained due to
geographical location and budget constraints. 

We provided them with an editing interface that looked and behaved very
similar to the e-learning player, but with added information needed to
be able to edit the information. Since the authors do understand the
structure of questions, this interface felt very natural for them. The
resulted content was also rich enough to publish the printed materials.

This project has also been very successful since they went live about a
year ago.

I think the question shouldn't be: "Why would I need semantic structured
content?" 
But rather: "What can I achieve when I do have semantic structured
content?"

Best,

Laurens van den Oever
CEO
Xopus Company

laurens@x...
http://xopus.com
+31 70 4452345

Waldorpstraat 17G
2521 CA Den Haag
The Netherlands

KvK 27308787

This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the sender. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail.