RE: Time to take a leadership position on client-side APIs

I second the proposal to consider this work, and that it should not be
confined to the MWI but ensure that the mobile use cases are give equal
priority at least when it is considered outside MWI.

I also agree that we need to ensure alignment/non-conflict on the data
properties exposed through DCCI, with:
- those defined in the DDR + Core Vocabulary
- those supplemented via OMA's DPE extensions for dynamic properties,
and based upon the OMA UAProf properties
- additional static/dynamic properties defined via the DCO
- properties defined though additional parties

I think the approach taken in DCCI (data model independence) gives us
the freedom to ensure this alignment. It's a very important issue that
will be one of the discussion points at the upcoming OMA meeting in
Beijing, as part of a review of OMA's review of the Core Vocabulary.

The Trust Model is a separate question, specifically on how the DCCI
through DOM is access-controlled. I think this is one of the first
issues to be considered, and not as a follow-on work.

We also need to ensure alignment with other access methods to device
properties, e.g. JSR's (are there specific ones already addressing the
same broad scope as DCCI?). MIDP security should provide a template for
managing application access to the device properties.

Outside DCCI / MIDP, as far as I know there are no other published API
methods for device property access. The OMA DPE Client access to the
device properties internally is implementation-specific. It would be
good also for use to consider guiding recommendations for
implementation-specific approaches.

Best regards,
Bryan Sullivan | AT&T 
-----Original Message-----
From: public-ddwg-request@w3.org [mailto:public-ddwg-request@w3.org] On
Behalf Of Dave Raggett
Sent: Wednesday, February 06, 2008 9:01 AM
To: Rotan Hanrahan
Cc: public-ddwg@w3.org
Subject: RE: Time to take a leadership position on client-side APIs


I agree with Rotan that this shouldn't be restricted to mobile devices,
but it is a question of how to build critical mass to make this happen
in a timely way, and right now, mobile devices have the spotlight.

  Dave Raggett <dsr@w3.org> http://www.w3.org/People/Raggett

On Wed, 6 Feb 2008, Rotan Hanrahan wrote:

> I agree with Dave's call to action, but I don't think that it is the
responsibility of the Mobile Web Initiative to create the roadmap, even
though the dominant use cases are in the mobile domain. The issue of
client-side APIs accessible to applications within browsers is not
confined to mobile. This is an issue that will touch all aspects of the
Web, and while I would accept the MWI being the initial champion of an
initiative on client-side APIs, it seems to me that the cause should be
taken up at a higher level within the W3C. Confining this work to mobile
would be a mistake.
>
>
> ---Rotan.
>
>
> ________________________________
>
> From: public-ddwg-request@w3.org on behalf of Dave Raggett
> Sent: Wed 06/02/2008 16:35
> To: public-ddwg@w3.org
> Subject: Time to take a leadership position on client-side APIs
>
>
>
> There is huge potential for mobile web applications that can access 
> device capabilities from client-side scripts. There has been a lot of 
> work on J2ME APIs, but we lack standards for exposing local device 
> capabilities to applications running in web browsers. The time has 
> surely come for W3C to bring interested parties together to work on 
> fixing this as a matter of priority.
>
> Properties like location, with privacy and associated legal issues, 
> will clearly be more complicated to deal with, as we will need to 
> address the security and trust models involved. But other properties 
> like battery level, signal strength, light and vibration control, 
> should be much easier to progress.
>
> The Device Description WG is defining APIs for access to properties of

> classes of devices, and the OMA is defining a protocol and server-side

> API for access to dynamic properties (DPE) that will enable servers to

> dynamically adapt media streams to match device orientation and 
> bandwidth. The UWA WG has recently moved DCCI to CR and published the 
> first draft WD for an ontology for the delivery context, where the 
> ontology is decoupled from the APIs that it models. DCCI is a 
> client-side framework, but doesn't itself define any properties. With 
> a little work, DCCI could be used for:
>
>     * dynamic content adaptation on client
>     * checking battery level, signal strength
>     * controlling the display brightness
>     * turning the phone's vibrator on and off
>     * checking screen orientation and size
>     * checking available free memory
>
> The following will need work on trust models and could be part of a 
> second wave:
>
>     * implementing location-based services
>     * interface to on-phone applications (PIM)
>       including calendar and contacts
>     * allowing web page scripts to initiate phone calls
>
> It seems timely for the W3C Mobile Web Initiative to create a roadmap 
> for building concensus on client-side access to device capabilities. 
> This seems like something W3C should be taking a leadership position 
> on given the opportunities for third party developers to stimulate 
> mobile data traffic if we succeed in standardizing the APIs. Without 
> such action there is a risk of fragmentation as multiple APIs appear 
> and developers have to choose between them.
>
> What's the best way to bring together the relevant stake holders?
> For instance, browser vendors, device vendors, network operators and 
> application developers?
>
>  Dave Raggett <dsr@w3.org> http://www.w3.org/People/Raggett
>
>
>

Received on Wednesday, 6 February 2008 19:05:06 UTC