Meeting Summary - 25 June 2007
[Weekly conference call, 25 June 2007] F2F well subscribed. TP topics. Normative names? Vocab contributions. Two docs to be published soon. Problem with tools for mapping IDL to languages. Details follow:
[F2F] We expect 10+ members to attend the London F2F next July. An agenda is being determined this week.
[TP Topics] The W3C is holding a Technical Plenary meeting in Boston in November and groups are being asked to consider topics for discussion. The harmonization of markup languages (XHTML-MP/Basic) is one possibility. Another is the (probable) false impression that having one markup language for all browsers is enough to ensure the success of the mobile Web. Other W3C groups need to be educated on the findings of DIWG/UWA, BPWG and DDWG in this regard.
[Names] There was a proposed resolution put to the group that only the identifier names in the ontology would be normative. Such names could be used (as strings) in API calls to identify the properties being stored/requested. However, the ontology might contain alternatives that could be useful for convenience in various programming languages (e.g. a camel-case version, an underscore_separated version, etc.). However, it was noted that other vocabularies could operate concurrently with the DDR Core Vocabulary, and the group has yet to consider issues such as namespacing to avoid confusion between vocabularies. Would the name include a namespace? Would the namespace be a separate parameter? Is there an alternative to namespacing? It was decided that the group would keep the proposed resolution on the table and wait until a better understanding of the naming needs of the API was achieved. (Generally, however, there was much sympathy for having normative names as part of the specification.)
[Vocab] The public vocabulary contribution process is active, as is the discussion mailing list. Everyone is encouraged to contribute. Rhys and Kevin are already preparing new material, and others are expected to follow.
[Publications] There were few wiki updates this week, but two of the legacy documents (Landscape and Ecosystem) are now substantially complete. The group is aiming to agree to publish these as final versions at the F2F. Some extraction from wiki to XMLSpec will be necessary as part of the publication process. Everyone is encouraged to do one last proof-read of the texts.
[IDL Tools] The group has identified a problem regarding tools to map IDL to implementation languages. Unfortunately, the tools that W3C has used in the past are no longer supported, and cannot be made to work properly. Nacho proposed that an XML version of IDL be used, from which various language mappings could be obtained via XSLT. Such a tool would be useful to the group to see how the API would appear in different programming languages, without having to develop/maintain these by hand. It was also proposed that the group use Java as a sample target programming language, and possibly even design/prototype via this language, though keeping the IDL as the normative definition. The question of how to determine if an implementation was conformant to the final IDL was also discussed. As the mappings are not unique (i.e. alternative mappings from IDL to Language-X are possible), it was suggested that black-box (functional) behaviour might be the only means to determine conformance, rather than inspection of the mappings.
[New Actions] (ACTION-51) Kevin to send e-mail reminder to group to review requirements doc.
Jose Manuel Cantera Fonseca (Telefonica)
Rafael Casero Escamilla (SATEC)
Dimitar Denev (Fraunhofer FIT)
Anders Ekstrand (Drutt)
Rodrigo Garcia (CTIC)
Rotan Hanrahan (MobileAware)
Martin Jones (Volantis)
Nacho Marin (CTIC)
Eman Nkeze (Boeing)
Jo Rabin (dotMobi)
Mike Smith (W3C)
Andrea Trasatti (dotMobi/WURFL)