These notes summarize the DocBook Technical Committee (TC) meeting that
took place in Washington D.C. on 6 and 7 Dec 2000 as part of the OASIS meetings
colocated with XML 2000. Meeting called to order by the chair. Moved: we will meet face to face for 1 day at XML Europe (Berlin)
and XML 2001 (Orlando, FL). We will request that our meetings be scheduled
on the last day of the conference so as to overlap with other technical
activities as little as possible. Furthermore, we will have telcons on
23 Feb, 27 Jul, and 28 Sep. The motion was adopted without objection. In accordance with earlier plans, DocBook 5.0 will be an XML DTD.
However, there was little support for completely abandoning SGML. Many
existing SGML systems will continue to be SGML based for some time and
there are features (most notably exclusions) that are lost in XML. Moved: the committee will produce an XML version of DocBook 5.0 with
sufficient parameterization to allow SGML features to be "turned on" with
a very small customization layer. The committee will provide the customization
layer. The motion was adopted without objection. There was some discussion of using the public text display version
to distinguish between the XML and SGML DTDs but it was eventually
rejected. Moved: the FPI for the XML version of DocBook V5.0 will be: and the FPI for the SGML version will be: The motion was adopted without objection. It was observed that the RFE database is in a rather jumbled state.
Action to norm: fix it. Moved: change the content model of SimpleMsgEntry
to allow MsgExplan+. The motion was adopted without objection. The fact that we were adopting a motion that made the content
model of SimpleMsgEntry more complex was not lost
on us. This continues to be an issue for Sun documentation. Recap: Authors sometimes need to provide a paragraph or two
of description before the items of a list. These paragraphs are considered
to be part of the list, so it's perhaps not appropriate to put them
before the list. Putting them before the list is problematic because grabbing
the list and moving it elsewhere will not capture the necessary introductory
material. Procedure already allows this structure.
No problems have been reported due to the fact that
procedure allows this structure. Moved: add %%component mix;*;
(minus the list elements) to orderedlist,
itemizedlist, calloutlist, and
variablelist before the first list item. The motion was adopted without objection. Tabled: waiting for feedback from Sun on whether or not
this solves the particular problem they've encountered.
Tabled: wait for the W3C XForms WG to finish then consider plugging it
in similar to the MathML plugin. Tabled: leave as customization layer until
interoperability is demonstrated. Action to the people who claim to
need this: give us feedback! Action: discuss on TC mailing list. Is this in scope? Do we
want to do this? Action: definitely in scope. Discuss on TC list:
do we add this as an element (like email) or a class on systemitem? The request for database-related elements seems already
to be handled by the database element. Moved: add a new class value to filename,
“extension”, to identify filename extensions. The motion was adopted without objection. Tabled: we don't understand the request. Action to norm:
ask for additional detail. Tabled: callouts in the middle of a paragraph seem odd.
Action to norm: ask for additional detail.
The motion was adopted without objection. The motion was adopted without objection. This would allow the contribution made by persons
given other credit to be identified. Discuss on TC list. What about
multiple authors? Thinking about XLink: Link/Ulink we think will be collapsed into Link using
XLink “simple link” semantics. Xref will remain because of the gentext processing
expectation. An empty "Link" cannot have the XRef semantic because it
conflicts with the empty "Ulink" semantic. Olink could be collapsed into link because xpointer
can replace targetdocent. Sun makes heavy use of OLink, action to Dennis: investigate the
problems that removing OLink would cause. Action to Norm: After Dennis reports, make a concrete
proposal before 5.0 is officially published (if XLink is a
Recommendation in time). Include an extendedlink element for use
inside *info elements. |
