RE: [dita] Issue for unifying acronyms and glossary

[Erik Hennum <ehennum@us.ibm.com>]

Hi, Terminology Folk:

I've updated the integrated terminology proposal:

http://www.oasis-open.org/apps/org/workgroup/dita/download.php/26657/IssueGlossary12026.dita

and, for convenience, the browsable HTML version:

http://www.oasis-open.org/apps/org/workgroup/dita/download.php/26656/IssueGlossary12026.html

This version makes the fixes proposed yesterday for the problems identified by JoAnn and Gershon on Monday (many thanks) including:

• Makes <glossdef> optional.
• Makes <glossPartOfSpeech> optional (assumed to default to noun if not specified).
• Moves <glossSurfaceForm> to <glossBody> and specialize from <p> on the grounds that the surface form will never have usage or linguistic properties; ie, it's better to treat the surface form as a property of the preferred form of the term rather than as an alternate form.

This version also fixes some expression issues noticed by Kara (including rewordings and corrections; again, many thanks).

I'd request that members of the Translation Subcommittee check for requirements that dropped out of this integrated proposal inadvertantly. In particular, please take a close look at the section on "Translation Issues for Abbreviated Forms" and consider whether the specified approach enables translation workbench software as expected.

Finally, given that some members of the Translation Subcommittee weren't at the full Technical Committee meeting on December 11th and may be wondering why an integration proposal even exists, let me summarize the problem that surfaced at that meeting.

The acronym proposal as submitted by the Translation Subcommittee supports markup like the following:

<abbreviated-form id="abs">
<expanded>Anti-lock Braking System</expanded>
<short>ABS</short>
<surface-form>Anti-lock Braking System (ABS)</surface-form>
</abbreviated-form>

The glossary proposal as approved by the Technical Committee in November supports markup like the following:

<glossentry id="abs">
<glossAcronym>ABS</glossAcronym>
<glossdef>A brake technology that minimizes skids.</glossdef>
<glossBody>
<glossPartOfSpeech value="noun"/>
<glossAlt>
<glossFullForm>Anti-lock Braking System</glossFullForm>
</glossAlt>
</glossBody>
</glossentry>

Obviously, the two markups provide an overlapping set of properties about the same thing:

• Both markups supply a referenceable identifier for the term.
• Both markups identify the acronym form of the term.
• Both markups identify the full form of the term.

If we can integrate the information from both designs, we can eliminate this redundancy problem.  DITA adopters will be able to declare a term once and have all terminology processes (including those we haven't anticipated yet) operate on that single declaration.  

For instance, here's the superset of the two previous examples under the current integrated proposal:

<glossentry id="abs">
<glossAcronym>ABS</glossAcronym>
<glossdef>A brake technology that minimizes skids.</glossdef>
<glossBody>
<glossPartOfSpeech value="noun"/>
<glossSurfaceForm>Anti-lock Braking System (ABS)</glossSurfaceForm>
<glossAlt>
<glossFullForm>Anti-lock Braking System</glossFullForm>
</glossAlt>
</glossBody>
</glossentry>

Adopters who need all three requirements can use the full set of properties.  Adopters with an exclusive goal of

• Glossary publishing can omit the entire <glossBody>
• Term resolution can omit the <glossdef> and <glossPartOfSpeech>
• Termbase population can omit the <glossSurfaceForm>

FWIW, I won't be able to do much more on this issue until the first week of January.


Thanks for taking a look and best wishes for the ending and new year,


Erik Hennum
ehennum@us.ibm.com