Minutes from call 2007/07/13

In attendance:

robert anderson (dita)

nancy harrison (docbook/dita)

rob weir (odf)

dick hamilton (docbook)

don day (dita)

michael priestley (dita)

mary macrae (oasis)

dee schur (oasis)

- michael gave an update on erik hennum's
work on a dita/docbook interop proof of concept - support for a dita map
pointing at docbook sections, transforming them to dita topics as part
of the build process; the flip side of the earlier proof of concept that
lets docbook embed a dita map and process the referenced topics into the
docbook publishing stream. work in progress.

- michael suggested that there is probably
a need/opportunity to preserve the source language's semantics in the target
format, perhaps using outputclass attribute in dita, or equivalent in docbook
(role attribute?) 

- dick suggested that a dita specalization
to provide an exact semantic match for docbook elements would be even better
than having outputclass 

- don suggested that a domain specialization
might be the best way to go, since it allows preservation of the meaning
of an element without trying to create a full docbook structure in dita
(which would be a force fit) 

- michael agreed that a domain specialization
seemed a good direction, while wanting to be careful of overlap/redundancy
(ie create specializations when there is a real semantic distinction, not
just a naming difference like p vs. para)

summary of existing transform options/activities:

- dita2docbook - demo exists in toolkit

- docbook2dita using map - demo from
erik

- dita enablement for docbook - using
docbook domain specialization - dick to investigate/drive

- docbook2odf - need something - converter
exists?

- odf2docbook - openoffice has filter
as starting point

- dita2odf - ask odf metadata subcomittee
if they have interest in collaborating here

- odf2dita - "

- also should investigate embedding
dita in odf, could then process the compound document to docbook 

- michael priestley suggested that the
odf interop story might not be just about word processor documents, but
also about spreadsheets - a potential match for ditamap tables, as well
as tables in content.

- rob weir suggested that the ability
to generate docs from a spreadsheet - like javadoc for spreadsheets - would
be useful

- michael conveyed paul prescod's hope
for convergence on infrastucture/syntax between standards, as well as content
sharing. for example, having the same attributes in all standards to identify
translatibility etc. would be very useful for translation vendors

- rob weir suggested we identify intersections/opportunities
as first stage - longer term effort, will depend on where each standard
is in a particular area (hard to change established practices, but should
try to avoid unnecessary further divergence)

- some discussion about how to coordinate
shared technical aspects of our standard; all agreed that the current discussion
list could evolve into becoming a tc that provides such coordination/guidance
- better to do as tc because liaisions don't scale well as new standards
emerge.

- potential path is discussion list
-> joint TC -> full TC

- dick re strawman: any way to simplify?
suggested eliminating one of the DITA specializations

venues:

- dita europe:

   - michael will be there
as keynote, can make sure interop is on the agenda

- xml 2006: in boston

   - they are interested,
will investigate further

- open publish: baltimore - anyone from
the baltimore area? would be good to have presence at both if possible

That's it for today - thanks all for
a productive meeting, and I'm off on vacation now.

Have a good one,

Michael Priestley

IBM DITA Architect and Classification Schema PDT Lead



http://dita.xml.org/blog/25