Meeting summary - 9 July 2007

[Weekly conference call, 9 July 2007] F2F preparations. More vocab
contributions. Wiki-to-doc. Working with OMA. Recording bugs and
limitations in the vocabulary. Using DI docs. Requirements, conformance
and assessments. Details follow.

[F2F] The face-to-face meeting is next week in London. The agenda for
the two day event is now complete, though the order of topics still
needs final confirmation.

[Vocab] Two more contributions to the DDR core vocabulary were received
recently. These will now be transferred to the public wiki, and may be
discussed on the dedicated mailing list. Andrea has agreed to do the
wiki editing for these contributions. Meanwhile, the BPWG is currently
discussing some potential contributions to the core vocabulary. DDWG
looks forward to receiving these from BPWG.

[Wiki] Some documents on the wiki will need to be transformed to XMLSpec
in advance of formal publication. The technical challenge of achieving
this will be investigated by some group members.

[DDR/DPE] The relationship between the DDWG and the OMA DPE group was
discussed. Both groups are working on technologies to enable solutions
take advantage of device properties. DDWG is working primarily on access
to "a priori" information, whereas DPE is working on "real time"
information. (This is a loose approximation.) There is an opportunity to
avoid unnecessary incompatibilities between the two technologies, and
indeed there is already a forming agreement to consider a common
ontology. DDWG awaits further information from DPE in this regard.
Meanwhile, the scope of work of the two groups does not necessarily
ensure a complete solution for contextually adaptive technologies, the
respective charters do not make this a requirement, and the use cases
for the respective specifications are unbounded (thanks to the
imagination of "customers"). The best we might offer is an informal
statement of intent: to work seamlessly together. This will be a topic
of discussion at the face-to-face.

[Properties] There was a comment recently regarding the specificity of
properties. The example cited was "table support" in browsers. From
practical experience, it is often not enough merely to assume that
support for a particular markup language (that includes tables) is
enough to determine that the browser supports tables. There are often
limitations and bugs. These are important data for adaptation
technologies. How should these data be represented in the DDR? Should
there be a new property introduced for each limitation discovered? This
topic will also be considered at the face-to-face.

[DI] It was decided that in order to explain how the API would assist
content adaptation, it would be good to make reference to the previous
publications of the DIWG with respect to "DI Challenges" and "DI
Techniques". Furthermore, the DISelect specification (originally from
DIWG and now being managed by UWAWG (UbiWeb)) is designed to be
extensible, so it was agreed that when the DDR API is complete, some
effort should be made to extend DISelect to support the DDR API. This
would provide a simple adaptive technology based purely on W3C
specifications, that could be used in profiles with other markup
languages.

[Requirements] It was recently agreed between editors of Requirements
and API that the description of "API Conformance Requirements" should be
made part of the API document, and that the existing Requirements
document should focus on "Assessment Criteria" for judging the quality
(and expected behaviour) of a DDR instance. One of the assessment
criteria would have to be that the DDR instance meets the conformance
requirements listed in the API document.

[Attendees]
Rodrigo Garcia (CTIC)
Rotan Hanrahan (MobileAware)
Martin Jones (Volantis)
Rhys Lewis (Volantis)
Jo Rabin (dotMobi)
Mike Smith (W3C)
Andrea Trasatti (dotMobi)

Received on Thursday, 12 July 2007 14:29:14 UTC