Category: Meeting Summaries

Friday, December 14th 2007

Permalink 23:45:47, Categories: Meeting Summaries

Meeting Summary - 10 Dec 2007

[Meeting Summary - 10 Dec 2007] Early start, Requirements and Vocabulary to be published, Call for API work, meeting schedule. No problems were reported with the early start time of the weekly call. [Requirements] Following an update to the Requirements document on the wiki, the group has resolved that the text of these wiki pages shall be recast in formal W3C style and published as soon as possible. The new document represents a high-level view of the DDR requirements, showing via use cases how these requirements are motivated. [Vocabulary] The most recent editor's draft of the Core Vocabulary, incorporating editor's notes, has been accepted by the group as a good initial publication and the group has resolved that it shall be published as soon as possible. [API work] With the work on the document publications decided, the group returns its focus to the development of the API. This work is currently being conducted via Java as a concrete design environment. To this end, sample code has been stored in public space within W3C, and this code is also being in an Open Source external environment. Members and observers will be encouraged to work with this code. Unfortunately, the group does not feel that the material is in a sufficiently stable state to justify publishing a formal document describing the work. Instead, the effort shall be directed towards technical development, with most of the work being conducted via the public mailing list. The next task in this work is to address the nature of the Property class. [Schedule] The last call of 2007 shall be Monday 17 December. The first call of 2008 shall be January 7. There is a publication deadline of 19 December, by which the group aims to have the Requirements and Vocabulary documents ready for publication. Attendees: Martin, Rotan, Matt, Dimitar, Rafa, Jose Cantera, Anders, Rodrigo, Nacho, Jo.
Rotan Hanrahan

Thursday, December 6th 2007

Permalink 09:54:30, Categories: Meeting Summaries

Meeting Summary - 3 Dec 2007

[DDWG Teleconference summary – 3 Dec 2007] Cross-referencing the ontology, populating the ontology, property names, XSD data types, Requirements, implementations, weekly calls [Cross-ref] The DDWG vocabulary needs to reference the UWA ontology for definitions of concepts and data types. However, a recent decision of the group was to refer to device information via "properties within aspects". This means that the mapping of DD properties to the ontology is no longer one-to-one. For example, a Vendor property can be applied to the hardware aspect (thus identifying the manufacturer) or to the browser aspect (thus identifying the origin of the software). The ontology has two different properties for these two vendor concepts. One approach to mapping DD items to corresponding items in the UWA ontology is to create a table that shows how the combination of DD properties and DD aspects map to the UWA ontology properties. It was proposed that this mapping could be part of the Vocabulary document. Three parts are identified: the DD properties, the aspects and the mapping to the ontology. Resolving these details could take some time, but meanwhile there is an initial draft of the vocabulary that is getting close to publication. It was agreed that this document should continue towards publication, with editor's notes to highlight areas of ongoing discussion. [Populating] It was noted that the UWA Ontology has information in it already, although it is understood that the intention is that the ontology will evolve based on the needs of respective vocabularies. The DD group needs to take care not to let the initial (experimental) information in the ontology control the evolution of the DD vocabulary. Instead, the group will continue to identify properties and aspects as it deems necessary to support content adaptation, and these will then be mapped to the ontology. If there are omissions in the ontology, these can be resolved in due course. [Names] A minor issue of the names of properties was raised. With the introduction of aspects, it is important not to have aspects appearing in the property names, For example, browserVersion would not be a good DD property name because it includes the browser aspect in its name. One example that occurred in early drafts was screenColorDepth. This was recently changed to displayColorDepth. The intention here was to replace the mention of the screen aspect with a term to describe how colour depth is transmitted as output. However, the word "display" could be seen as synonymous with "screen" so this might need to be reconsidered. [XSD] In early drafts of the vocabulary, data types were mentioned as String, Boolean, Integer etc. The group resolved that these data types should be formalised by reference to the XML Schema Data types (XSD). [Requirements] A recent update of the requirements document on the wiki was noted, and commended. Members will review the text, with the intention of agreeing to publish. The new version uses use cases to explain the motivations behind certain DDR characteristics. [Implementations] It was noted that there is growing Open Source activity in support of the DDR API. It is expected that during 2008 there will be free, commercial and Open Source implementations of the API. Announcements of these have been made on the public mailing list. [Weekly call] With some participants attending the weekly call very late at night in their time zone, it was resolved to move the call time forward an hour from this point onwards. Attendees: Rotan, Jo, Dimitar, JongPil, Rafa, Matt, Pontus, Jose Cantera, Martin, Rodrigo.
Rotan Hanrahan

Sunday, November 25th 2007

Permalink 15:07:57, Categories: Meeting Summaries

DDWG F2F summary published

A summary of the November DDWG Face-to-Face meeting was reported recently on the public mailing list: There was considerable progress during the meeting, and some drafts of the DDR API interface were developed. Publications are expected soon. Read the report for more details.
Rotan Hanrahan

Wednesday, October 24th 2007

Permalink 22:09:47, Categories: Meeting Summaries

Meeting Summary - 22 Oct 2007

[Meeting summary - 22 October 2007] End of properties, Korea agreed, Boston scheduling, Publications. [Properties] The wiki now reflects the state of the group regarding decisions about properties that are to be part of the Core Vocabulary. It was decided that next Monday would be the end of this round of deliberations on the vocabulary, with all future discussions to take place from January. This will give space to the group to concentrate on the API. One of the issues not yet resolved is the significance of device "brand" and "model". As the association of hardware/software to brand/model is not one-to-one, is unpredictable and is not generally not essential for real time adaptation, it's value as a core property is questionable. However, for non-realtime authoring phase use, the brand/model is very useful. It was suggested that perhaps the group's work on structures could provide a mechanism for associating brand/model information with entries in a repository. This remains an open issue. [Korea] Following a group vote, a decision was reached to accept the invitation to hold a face-to-face meeting in Korea next March. [Boston] An ongoing poll reveals that there will be several active DDWG participants at the Boston face-to-face in November, but (not surprisingly) several are also engaged in work with other W3C groups during that week. The group chairs will coordinate their respective agenda to make optimal use of member availability. [Publications] The two legacy documents are on track for publication. A draft of the Core Vocabulary document will be attempted during the F2F. With the discussions on the vocabulary being parked from next Monday, there should be an opportunity now to concentrate on the API document. Meanwhile the group also hopes to close the discussion on the outstanding high-level Requirements document during the F2F. [DOM bindings] A recent (Oct 17) draft document from the CSS Working Group includes several items of interest regarding the use of IDL for specifications involving DOMs. Some of this information can inform the DDWG's work on drafting an API for the DDR. An issue has been raised to ensure that members of the group become familiar with the document. [Admin] The group's internal tracker now has the means to flag items for review. There has also been some progress on the merger of the group wiki to a new platform. Attendees: Anders, Andrea, Bryan, Dimitar, Jo, Matt, Martin, Mike, Rafa, Rotan.
Rotan Hanrahan

Thursday, October 18th 2007

Permalink 15:54:45, Categories: Meeting Summaries

Meeting Summary - 15 October 2007

[Meeting Summary - 15 Oct 2007] Vocabulary resolutions, Korea F2F. [Vocab] This week’s conference call was energetic and productive. The main topic on the agenda was to make decisions on the many proposed core vocabulary properties, taking into account the many responses and other comments received to date. Over the past week, the group had been voting to identify those properties that were core, not core and deserving of further debate. It was assumed that those who had engaged in the voting had strong opinions to express. The first property discussed related to H.264 (Baseline Profile Level 1.0). This deals with streaming video rates. It was felt that this property was outside of the scope of a core vocabulary that was focussed only on the essential properties for adapting Web content for mobile delivery contexts. The group intends to respond to the proposal by indicating how domain specific, custom and advanced vocabularies can be created; and the role of the UWA ontology in this regard. It was resolved that H.264 is not part of the core vocabulary. The next property was the DOM Level of the browser. It was felt that the DOM Level did not have a strong bearing on the adaptation of content. Knowing the DOM Level might be useful in adapting scripts that access the DOM. In the same vein, the group considered the XMLHTTPRequest support of the browser, and again this was felt to be a script-oriented property. Both of these properties would be considered core to an adaptation process for mobile Ajax, but not for a simple adaptation process that is dealing with simple Web content. It was decided that these properties would be communicated to the OpenAjax Alliance, with a view to encouraging them to create a vocabulary for adaptation of Ajax, following the example of DDWG. Support for ECMAScript (or JavaScript) was also discussed. It was felt that knowing that a browser supports scripting is beneficial for more than just Ajax applications. It was therefore resolved that this property should be in the core vocabulary, but there would still need to be further discussion on whether it should represent ECMAScript and/or JavaScript. The proposed property of Characters Per Line was debated. It was suggested that this property would be useful for pagination. However, it was also seen as a throwback to the old days of WML content, where fonts were very simple and the character density was an important property. Today, with variable fonts and more complex layouts, character density is less precise. In the absence of exact font information, and approximation based on screen resolution would probably suffice. To do things like sophisticated line wrapping, you would need much more detailed information about the fonts, which the group agreed would go far beyond a simple adaptation. For simple adaptation, the screen width would be a sufficient means to approximate the character density and thus the Characters Per Line property was resolved to be non-core. It was pointed out that no matter what the DDWG decides will be in the core, there will be people who will say some properties are missing. The best that the group can achieve is a basic vocabulary that will support simple content adaptation, and be sufficient to demonstrate the operation of the DDR API. This needs to be explained in the Vocabulary document. The group then turned its focus on some properties that all agreed were core: the dimensions of the screen. The width and height were seen by everyone as essential for content adaptation. While previous proposed properties were rejected from the core, these two properties were immediately resolved to be core properties. With this practice on the decision process, the group agreed to take the discussion to the electronic forum, where the remaining candidate properties would be debated, and some further resolutions taken. This debate, and the conclusions, would be visible on the public mailing list and the wiki. [Korea] Finally the group acknowledged an invitation to hold a face-to-face meeting in Korea in 2008. The group will send a formal response after polling the members. Attendees: Matt, Andrea, Pontus, José, Jo, Rotan, Martin, Dimitar, Kevin, Jongpil, Bryan, Nacho
Rotan Hanrahan

<< Previous Page :: Next Page >>