Meeting Summary - 18 June 2007
[Weekly conference call, 18 June 2007] Welcome Matt. Top N to be contributed by group. Ecosystem updated on wiki. Issues discussed: ontology annotations, property aliases, static/dynamic terminology, DOMString/wstring/etc in IDL. Taskforce report. Details follow:
[Matt] Matt Womer joins the DDWG as the W3C's staff contact, and is already clocking up action items.
[Vocab] Volunteers from the DD group will add entries to the vocabulary this week, based on suggestions derived from the Top N properties cited in charter 1. To avoid overlaps, the chair shall distribute suggested entries. The form will be updated to draw submitters' attention to the ontology in progress, so that new entries can be devised in light of existing entries.
[Ecosystem] The Ecosystem document is near completion. Some final inputs to the Trust section and References will complete the document and it will then be ready for consideration by the group for publication as a Working Group Note.
[Annotations] The issue of meta-metadata was considered. For example, the origin of contributed vocabulary entries, or the origin of the values for such entries. It was concluded that such information does not directly serve the needs of content adaptation and therefore is probably best served as an administrative feature. Such information could be used by an administrator to determine the trust to associate with different sources of data. The ontology may have more data (and types of data) than are directly accessibly by the eventual DDR API.
[Aliases] The group RESOLVED that alternative names for properties (e.g. as might be used in implementations in different programming languages) would not be represented via sameAs or equivalentProperty mechanisms. Instead the group will use the current annotation-based mechanism (i.e. a "see also" reference). This means that reasoning systems are restricted to the normative property identifiers, and will be unaware of the alternative names.
[Static/Dynamic] The problem of terminology (static vs dynamic) was again raised. As an example, the group considered the screen height as a property which normally has a fixed value, but for some devices the screen can be adjusted (e.g. via a flip-out component) and therefore it has more than one value. In such cases a default value may be useful if you cannot determine the current value. Meanwhile, on the client side, this is a value that can be read directly without recourse to a device description repository. Such properties blur the distinction between static and dynamic. It was decided that a wiki page to explore these issues would be created.
[Strings] The issue of string representation in IDL was considered. Unlike the W3C DOM-related IDLs, the DDR API is a server-side API that has no immediately obvious need for DOMString. Nevertheless, the normal IDL "string" may not be sufficient to represent certain values that may be returned by the API. Therefore some consideration must be given to the options that are available, including string, DOMString, wstring and some custom definition that the DDWG may create itself. Furthermore, the issue of character encoding for wstring types was also considered (and seen to introduce more questions). A call was made to find more authoritative information on the use of strings in IDL.
[Taskforce] The new taskforce met last week and there are some draft conclusions. It is suggested that the context key should be opaque. An approved version of the conclusions will be circulated soon for comments. The results of the taskforce are on schedule for delivery to the F2F meeting scheduled for mid-July.
[New Actions] (ACTION-49) Matt to set up registration for July F2F. (ACTION-51) Rotan to start a Static vs Dynamic wiki entry.
Pontus Carlsson (Drutt)
Rafael Casero Escamilla (SATEC)
Dimitar Denev (Fraunhofer FIT)
Rodrigo Garcia (CTIC)
Rotan Hanrahan (MobileAware)
Martin Jones (Volantis)
Rhys Lewis (Volantis)
Nacho Marin (CTIC)
Jo Rabin (dotMobi)
Mike Smith (W3C)
Andrea Trasatti (WURFL)
Matt Womer (W3C)