OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

dita message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Q&A from DITA Briefing #1 (Intro to DITA) -- 5/25/2004

DITA Briefing #1, "Intro to DITA" -- 2004-05-25

Questions and Answers

    DD: indicates Don Day
    MP: indicates Michael Priestley

Sharon -- 

    - Question about titles -- are they part of the content or the

            - MP: Titles are content of topics, but DITA topic
              content also provides <titlealts> which includes
              navtitle and searchtitle, which are effectively
              metadata in that they are optionally available for
              application roles that the content title is less
              suitable for.  An example might be a shorter navtitle
              to fit a known width constraint in a left-hand
              navigation pane, or a search title that includes
              a product name in order to match expected search

Scott Tsao -- 

    - Reuse of design, reuse of content, reuse of processing -- just
      wanted a clarification of what that means.  His understanding
      was already correct.

    - Does DITA restrict you to XSLT for processing?

            - MP: No, but Michael does like it; it fits well with
              DITA.  DITA was originally designed to work
              specifically with the W3C-based processing
              architecture, including CSS, XSLT, XSL-FO, XLink, and
              so forth.  For your topics to be interoperable with
              other DITA users, any processing based on other tools
              should nevertheless have similar behaviors for
              specialization (class-based lookup, safe fallthrough
              to base classes, conref resolution, etc.)

    - When you talk about reuse of design, content, and processing,
      does that mean reuse of Schema components, of instance
      content, and of XSLT routines? 
            - MP: Yes to all.

    - Confused when MP says he wants to use the map to capture the
      metadata, whereas the topic itself doesn't need to carry the

            - MP:  You can author the metadata in either place.  You
              get the most reuse possibilities when it's outside the
              topic.  It is designed to be additive.

    - If I have a set of components I want to standardize, and one
      of them is reused in many places and I want to store it in
      a library, would I create a topic for the definition?

            - MP: This is more of a best practice question: how
              should new specializations make use of the various
              levels of design reuse and content reuse in DITA?  For
              example, in a glossary application, should the new
              design be based on definition list or on nested
              concepts?  Either would work.  Michael favors reuse at
              the topic level for this, but the right design will
              probably depend on other design factors that will
              differ between communities.  DITA's features (map,
              topic, conref, etc.) enable a wide number of
              information architectures to be incorporated;
              community agreement will be the most important factor
              in defining how/whether that IA interoperates with
              other information architectures.

              Wading a bit deeper here, the value of a topic-based
              set of definitions is that the topic specializations
              can be very specific, providing support for etymology,
              pronunciation, synonums, usage, etc..  Upon output,
              only the basic term/def can be output to a specialized
              definition list that serves as an output-specific view
              of the separately maintained topic-based definitions.
              There are many cases of use-driven differences in how
              common structures might be supported in
              specializations.  Call-center issues (topic level) can
              be abstracted into QandA lists (FAQs), message
              definitions (topic level) can be abstracted into
              message sets for conversion into resource bundles,

Padma Neppalli (Intel) --

    - How do you refer to something that's buried somewhere inside
      a topic?

            - MP: Most DITA elements have an ID attribute and
              a conref attribute. You can reuse content from an
              element with an ID by pointing to it from another
              element of the same type with a conref attribute.

            - DD: Reuse is not only topic-to-topic but also within
              a topic.

    - When are conrefs validated?

            - MP:  At build-time, when you create the output

              Also note: "valid" means more than just "is the
              content ok to copy over" - it checks whether the rules
              are the same in both places - i.e., you can only reuse
              a paragraph from someplace where paragraphs are valid,
              and you can only reuse paragraphs that have the same
              element content rules. This will be explained in more
              depth in a later presentation.

            - DD: Editing tools often will show you broken links
              also. It can tell you whether or not you are pointing
              to a valid conref, and helps you select only valid
              content.  Similar to ID/IDREF which does not work well
              in a cross-document architecture like DITA.

Eric from BMC Software --

        - How does DITA create the equivalent of an index?
                - DD:  DITA topics contain an index term element.
                  They can be nested for creating hierarchical

                - MP:  The public tool doesn't include indexing
                  tools, but the markup is there.  MP thinks it's
                  better to do the indexing on the map level -- it's
                  easier to control index design.

                - DD:  But if you put the index data into individual
                  elements, they can be picked up wherever those
                  elements are conref'd.

                - MP: This means you can't get rid of them when you
                  need to. Main point is that indexing is something
                  you do in terms of a collection of topics - your
                  choice of first and second level headings is
                  determined by the size and subjects of your
                  topics. If these choices are embedded in your
                  content, then your content will create bad index
                  entries when it is reused in a new context.

                  IBM still relies upon their SGML indexing
                  tools -- they convert backwards to SGML whenever
                  they need an index, and use those tools to create
                  an index.  Obviously this isn't helpful to anyone
                  else, and MP is hoping that DITA-based transforms
                  for indexing will be available at some point.
                  There *is* a full indexing ability, just not fully

                - Elliot Kimber:  Index processing that's part of
                  the DocBook XSLT can probably be ported to DITA
                  with minimal effort.

                - DD:  Reusing legacy processes -- If you already
                  have a legacy processor, you can port DITA
                  backwards to it so you don't lose that tool.  Thus
                  what MP presented as a potential shortcoming might
                  be a benefit for people who have useful legacy
                  tools already working.

                - MP:  Good point for the general rule about legacy
                  compatibility. That said, indexing is a likely
                  candidate for standard processing.

Scott Tsao -- 

        - Does anyone know about UDEF (a particular standard used by
          airlines) and a possible intersection/overlap with DITA?

                - DD:  Eric knows about this -- let's take
                  this to the mailing list.

Mary Parmelee from SAS Institute -- 

        - Are non-TC members invited to subsequent briefings?

                - DD:  Yes.  Go to the TC website for current
                  information.  Just let Don know if there will be
                  a need for him to reserve additional lines, and
                  try to call in from local conference rooms if
                  possible to conserve lines.

No more questions from the presentation.  Keep an eye on the website
for any changing news.

Seraphim Larsen                  ICG Technical Publications
Sr. Technical Writer                      Intel Corporation
(480) 552-6504                                 Chandler, AZ

The content of this message is my personal opinion only. 
Although I am an employee of Intel, the statements I make 
here in no way represent Intel's position on the issue, nor 
am I authorized to speak on behalf of Intel on this matter.

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]