RE: [dita] Mixed content in DITA

[Michael Priestley <mpriestl@ca.ibm.com>]

The <task> topic type provides
good examples here - for example, <context> is a specialization of
section that has no title, and often contains no more than a single sentence,
although it can contain more, for example a couple of paragraphs.

Michael Priestley
mpriestl@ca.ibm.com

"Paul Prescod" <paul.prescod@blastradius.com>

04/09/2005 09:00 AM

To	"Don Day" <dond@us.ibm.com>
cc	<dita@lists.oasis-open.org>
Subject	RE: [dita] Mixed content in DITA

I can understand why you want generic elements to be quite
loose so as to allow tightening up in specialization. But obviously there
are limits otherwise the content model for everything in DITA would be
ANY. I'm not clear either why it is valuable to

a) have sections (or specializations of sections) without titles
b) have sections (or specilizations) with text content that is NOT contained
by paragraphs

Perhaps the goal is to emulate HTML's very generic "DIV" element
which has almost no semantics. If so, I would propose that what we now
call section should be a sort of abstract base element (similar to what
was described for keyword) and that another element should be provided
for the kind of sections that authors typically use. Yes, each particular
customer can do this themselves in specialization but we are finding that
some smaller customers prefer to use DITA out of the box as they do Docbook.

 Paul Prescod