Archives for: May 2007
Monday, May 28th 2007
Meeting Summary - 28 May 2007
[Weekly conference call, 28 May 2007.] Next F2F in July, Taskforce commencing, Wiki updates (IDL, Ecosystem), IDL tools. Details follow:
[F2F] Following a poll of the group members, the majority of respondents agreed that July 16–17 was the right time to hold the summer face-to-face. We will be hosted by Vodafone, who have kindly offered a room and meeting services. Some participants will be missing, which is unavoidable during the vacation season.
[Taskforce] Three volunteers have indicated they will join the "Context Key Taskforce". A number between 3 and 5 is desirable. A wiki page has already been prepared to record the results of the taskforce. A deadline for a report has been set for the week before the July face-to-face.
[IDL] Nacho announced several updates to the wiki covering IDL issues. A W3C tool has been identified that can assist in the production of documented IDL material, but the tool is 5 years old and may be difficult to get working again. Volunteers will look into the code, and try to get help from W3C editors who have used the tool in the past. The tool, if we can get it to work properly, will help as the complexity of the IDL resources grow. Given the lack of guarantee of getting this tool to work, other possibilities have been suggested, such as using Java as a design language from which the IDL can be derived (namely because more people understand Java and tools are available to assist).
[Ecosystem] Several new updates have been added to the Ecosystem document. The updates can be seen marked in the Ecosystem wiki pages, and in the wiki diff tool. The "Business Model" section seems out of place, and as it was intended to deal with commercial matters, it might not be a topic with which the group can speak with authority. Therefore the section has been made a candidate for removal.
[Requirements] The requirements document has received major updates recently. Some checks will have to be made to ensure that the requirements agreed during charter 1 are still reflected in the updated charter.
[New Actions] (ACTION-45) Rotan to follow up on the possibility of the specification being made via Java-to-IDL.
Anders Ekstrand (Drutt)
Rafael Casero Escamilla (SATEC)
Jose Manuel Cantera Fonseca (Telefónica)
Rodrigo Garcia (CTIC)
Rotan Hanrahan (MobileAware)
Cedric Kiss (W3C)
Nacho Marín (CTIC)
Jo Rabin (dotMobi)
Mike Smith (W3C)
Monday, May 21st 2007
Meeting Summary - 21 May 2007
[Weekly conference call, 21 May 2007.] Vocabulary process to be active
this month. Taskforces to be set up. Seed IDL. F2F schedule. Details
[Vocabulary Process] Following a short assessment of the proposed Core
Vocabulary Process, a decision was taken to launch the process in
public. This means setting up a Web form via which anyone can
contribute, subject to having or obtaining a W3C account. If necessary,
the form will also have an email equivalent. Contributions will be
semi-formal. It will be the responsibility of the DDWG members to
process the contributions. The form will be accompanied by an example to
illustrate the process. Discussion of the contributions will take place
on a public email list. It is hoped that the process will be running
before the end of the month.
[November F2F] The group agreed to co-locate a face-to-face meeting with
the W3C Technical Plenary scheduled for November.
[Summer F2F] The timing of the summer face-to-face provides some
scheduling challenges. Following an examination of the options, a late
July meeting appears to be the leading preference, but we need to take a
poll of the group to be sure. Getting a venue will be the next
challenge. Given the decision attend the November TP (in Boston), a
European venue for summer is indicated.
[Group Status] The recently blogged group status was seen as a fair
reflection of the work to date. Group members were encouraged to take a
more active part in the group blog.
[Taskforce] In order to focus on particular issues, the Chair proposed
that one or more taskforces be established within the group. These would
comprise people who had particular expertise and experience relevant to
the task. The first taskforce to be set up will be one to examine the
issue of a "key" to represent the context (device and/or software) in
which data is stored or retrieved.
[Seed IDL] An initial IDL for a minimal repository was circulated on the
public list to foster discussion. The group is now consuming this new
material and is expected to respond later.
[Documents] Two of the legacy documents have not progressed recently,
but the Requirements document has moved since the recent F2F. New
material is being entered into the wiki this week. The editing is taking
place on the wiki in order to provide a public view of the process. The
final document will be derived from the wiki and represented in the
normal HTML format used for official W3C documents. The API document is
expected to progress once the IDL discussion gets under way.
[New Actions] (ACTION-40) Rotan to work with Cedric to get form set up
this week. (ACTION-41) Rotan to take e-mail poll about July f2f.
(ACTION-42) Rotan to contact DanA/Vodafone about possibility of hosting
next f2f at Vodafone. (ACTION-43) Rotan to initiate TF to investigate
resolving key issue. (ACTION-44) Kevin to update the Wiki pages with
notes he has take on Requirements doc.
Saturday, May 5th 2007
12:32:58, Categories: News
State of Play, May 2007
The DDWG maintains a wiki in public, has a public mailing list and (of course) this blog. But that is not all. The group also has regular teleconferences, some face-to-face meetings and there’s a fair amount of email traffic between the members outside of the normal public channels. I have pulled all these together to get an overall picture of the state of play in DDWG. What I find is this:
The group work commenced properly in January 2007, around the time of the first face-to-face. The Top N Properties information was to be published within 2 months, and was actually released via the wiki in February.
It was decided that IDL would be used to define the API and that the vocabulary should be developed through RDF/OWL. The vocabulary would need to be consistent would other technologies that deal with device characteristics that were in development, or being planned, by other groups. It was agreed that we needed a common ontology of device information. Work on the ontology commenced immediately.
The work proceeded from February until responsibility was transferred to the newly chartered UWA group. The ontology is currently being developed as a joint UWA/DDWG project. Meanwhile the DDWG has continued its work on definitions of terms, revising the requirements and other legacy documents. Even the meaning of "Device Description" became the topic of intense debate.
The work on DD Structures commenced in March. This deals with the issues of grouping device descriptions in a portable fashion, so that one can (for example) retrieve information about a device family, such as a PDA. Like all other DDWG work, you can view the details on the wiki.
One of the tricky issues being addressed by the DDWG at this time is the "identity" of a device. The charter requires the group to decouple the concept of the physical device from the software (user agent). This is important because in order to insert or retrieve information to/from the DDR it will be necessary to unambiguously identify the "device" to which the data is associated. Given the malleability of mobile hardware and the configurability of mobile software, devices have become a bit of a fuzzy concept. Consequently, a unique key to identify these things has become hugely complex. This issue consumed a whole day of debate at the April face-to-face meetings. In the end, we decided that the key would likely be a compound derived from contextual information, mainly something to identify the physical device together with something that identifies (characteristics) of the software. The DDWG might have to mandate a key generation algorithm to accompany the core vocabulary, though this will require much more debate, given that the DDR is supposed to be extensible.
The details of the keying mechanism have been set aside while the shape of the API is developed. This will be done in tandem with the completion of the requirements document, which is being recast to clarify its relationship to the DDR and the API.
Meanwhile the process of creating the core vocabulary has commenced. Behind the scenes, the group is testing a contribution process, to verify that it can gather the right information for the core vocabulary, and can feed into the UWA ontology. This will be tested for a little while, and then it will be launched in public, empty and waiting for contributions. Expect to see the first signs in May 2007.
As chair, I am pleased to see that the foundations for the API design and the core vocabulary are fitting into place, and happy to see that DDWG will soon be collecting the elements of the core vocabulary. The challenges of the API design, the device identity mechanism and the device description structures are going to be tough, but I believe we have the team to succeed.
Finally, I hope to see more public engagement in the DDWG work, so please pay regular visits to the wiki and subscribe to the mailing list. Your contributions are very welcome.
Rotan Hanrahan, Chair of DDWG.