Re: [dita] The whatever-we-call-it factor

["Dr. Stanley Doherty" <stan@modularwriting.com>]

Hi Bob --

I like your approach very very much. When I teach DITA, I emphasize that transformed output is not the "thing" by which we measure the success or integrity of a DITA implementation. Your "information network" thingie would give me a term for the DITA gestalt prior to it being transformed into anything. 

Hail "thingies"!! (or whatever we call it)

/stan   


I have done a preliminary reading of the Linking topic cluster and Eliot's email that we discussed last Tuesday. First, I believe that it's appropriate to address the facets of DITA linking in a central place. It is the linking that transforms a collection of resources into a useful DITA whatever-we-call-it. There is enough complexity that a unified discussion is needed to furnish context for the normative parts of the spec.

This "whatever-we-call-it" nomenclature is annoying, and I promise not use again. But it emerged because the spec is a headless description of DITA. It never really says what a DITA something-or-the-other is. One consequence of this shortcoming is that there is no foundation to furnish context when explaining DITA structure. 

When I was a technical writer one of the most effective ways I found for getting information from people was to just make something up, and then ask them to approve it. In that spirit, I offer this attempt give the DITA spec its head.

DITA is an XML vocabulary for creating content and then aggregating it and other resources into a structured information network. Information networks get created for some purpose. For example, an author could create a DITA information network that includes all of the content and structure necessary to publish a user guide for some product.

All relationships among resource entities that contribute to a DITA information network are expressed through a combination of XML-element containment (native structural relationships) and DITA linking-attribute values. An information-network’s native structure is primarily defined through the hierarchical arrangement of linking-elements contained within DITA maps. Other forms of DITA linking allow a variety of non-structural content relationships to be expressed.

Typically, a DITA information network starts within a specific DITA map. A DITA map that serves in this capacity is known as a root map. A DITA map that serves as a root map in one information network could also serve as an information subnetwork in another information network. However, such a DITA map would necessarily lose its status as a root map in that subordinate context because it would no longer be the entry point for that information network.

DITA information networks typically contain independent information subnetworks. These subnetworks are bound together through the use of linking elements within DITA maps. In some cases, these information subnetworks express useful relationships that do not contribute to an information network’s native structure. The following DITA constructs are examples of that: subject scheme, relationship tables, and key definitions.

An information network may be also be constrained through filtering. With filtering, one or more ditaval documents specify the filtering-conditions to be applied, and selection-attribute values within DITA topics and DITA maps bind filtering-conditions to content.

Best Regards,
-- 
Bob Thomas
+1 720 201 8260
Skype: bob.thomas.colorado
Instant messaging: Gmail chat (bob.thomas@tagsmiths.com) or Skype
Time zone: Mountain (GMT-7)