MHonArc v2.5.0b2 -->
dita message
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]
Subject: RE: [dita] Thoughts On Sections
As you rightly point out, topics are
extremely loose in content model - about the only thing we can claim, at
that level, is that they are individual topics. My own rule of thumb is
that if the subheading has more than one paragraph and has a unique title
rather than a repeating one ("Example" or "Usage" would
be repeating subheadings, in the sense that they are common across multiple
topics) then it should be a nested topic rather than a section. Topic-oriented
authoring is an existing methodology for technical communication and instructional
design, and DITA definitely has its roots there.This distinction between sections, which
are divisions of a single subject, and nested topics, which are new subjects
within the domain of a parent one, is core to the ability of DITA to manage
topics as individual units in maps. In fact, the focus on topics is as
core to DITA as specialization. Michael Priestley
IBM DITA Architect
SWG Classification Schema PDT Lead
mpriestl@ca.ibm.com
> Would it not be preferable to advise users to
consider
> developing these subconcept statements as individual topics
> that may then be more easily reused in other contexts? In
> most cases, the nesting into two or more levels of
> information is a formalism preferred by the writer rather
> than a useful construct for the reader. In most instances,
> when I review these structures, they are poorly structured. I
> know that we can't enforce good writing practices, but it's
> interesting to think we might try.
In this case, the structures are determined by the organization.
Nevertheless, their point of view is that the parts are inherently
related and not standalone topics. The readers expect the structure
exactly as it is. I don't see it as that rare a case for an organization
to have a template for an information type where the template has
sub-structure.
> If the authors were asked to write smaller conceptual topics,
> with better structuring of information to make the
> information more flat (and therefore perhaps more
> understandable), we could stay with the original design. If
> the nesting is really that important to understanding in the
> output context, the nesting could occur in the ditamap rather
> than in the topic itself. The smaller conceptual topics would
> also be more readily available for reuse in contexts where
> they should be standalone.
The nesting is not just important in the output context. The nesting is
how the domain experts (both creators and consumers of the information)
think about it. It is their data model and they are happy with it. Is
there any particular reason that DITA should disallow them from creating
a specialization that reflects this? DITA's base (pre-specialization)
topic type is so loose in most ways that it seems perverse to enforce
this one rule.
Paul Prescod
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]