RE: Comments on Device Description Repository Requirements 1.0

(For your benefit, I identify myself, my role and employer below, though
in this exchange I am only acting as the Chair of the DDWG.)

I don't believe I have proposed any particular solution in preference to
any other. The idea of an ISP acting as an adapting intermediary is a
solution I have recently suggested when you have no means to act
directly upon the original content. In some cases it may be possible to
adapt the content within the device itself (see Opera, for example),
though this requires a more capable device, and may result in more
content being transferred to the device than would otherwise be
necessary if some of the decisions could be made before the content was
delivered.

You identified a product in your email. Other vendors offer technologies
that have a somewhat similar approach. The technology builds upon
existing Web authoring solutions, mainly XHTML, plus some metadata and
author-determined decisions that can be exercised whenever the delivery
context requires. The product you identified is indeed an origin-server
adaptation solution. (There are other products that can act as
intermediaries.)

The W3C Device Independence Working Group is following a similar
strategy. The aim is to combine and build upon established W3C
technologies, and provide a mechanism for context-sensitive adaptation
(mainly selection based on simple expressions). This approach will also
support origin-server architectures, though other approaches will also
be possible.

Maintaining the part of the delivery context representing the static
device capabilities will be the responsibility of the people behind the
DDR. This is much like the responsibility for maintaining the mapping
between names and IP numbers (i.e. the DNS). A distributed
responsibility. There may be some central responsibility, but until we
discuss implementation in depth we will not know which is the best
approach. Some collections of information already exist (though they are
not offered as services). The OMA is collecting UAProf instances, and
the WURFL community regularly update their XML-based collection of
device details.

---Rotan.

____________________________
Dr Rotan Hanrahan
Chief Innovations Architect
Mobileaware Ltd
 
3094 Lake Drive
Citywest
Dublin 24, Ireland
E: rotan.hanrahan@mobileaware.com
W: www.MobileAware.com

 

-----Original Message-----
From: Sam Sotiropoulos [mailto:sam@sotiropoulos.com] 
Sent: 06 May 2006 17:32
To: public-ddwg@w3.org
Cc: Rotan Hanrahan
Subject: RE: Comments on Device Description Repository Requirements 1.0

Having taken a look at the mobileaware.com site which I assume is Mr.
Hanrahan's employer, I found the following under their Products menu
item:
--------------
"Mobile Interaction Server delivers Web-based content and business
applications to all mobile devices. Whether you are seeking to extend
existing enterprise applications to your customers, partners and mobile
workforce or looking to deliver advanced mobile data services to attract
more subscribers, the Mobile Interaction Server offers a proven solution
for delivering your business content to mobile devices.

Mobile Interaction Server seamlessly extends enterprise applications,
portals and web services allowing them to be accessed by any mobile
device at any time. It overcomes the challenges of enterprise mobility
by allowing businesses to:

    * Optimize mobile content delivery and presentation by providing
easy to use plug-ins to popular authoring tools for designing highly
personalized content and graphical user interfaces tailored for each
specific mobile device.
    * Rapidly develop and deploy multi-channel web applications by
integrating web and wireless services into a common framework.
    * Extend the capabilities of existing enterprise infrastructure by
enabling mobile device access to back-office applications and databases.
    * Cost-effectively extend customer and employee self-service web
portals to support mobile access.
    * Future-proof the investment in enterprise mobility by adopting
mobile technologies that conform to industry open standards and offer
complete network, protocol and device independence for maximum
flexibility and control."
-------------

Having already read the Device Description Landscape working draft
(http://www.w3.org/TR/dd-landscape/#sec-theneedfordevicedescriptions),
it appears to me that what Mr. Hanrahan is proposing can be likened to a
kind of subclass of ISP (in a loose sense) that will provide the highly
specialized service of device adaptation as an intermediary step in the
content origin, transmission and reception cycle of web delivery.

I have to admit, I am skeptical about the (real, perceived or projected)
need for such an approach.  Call me old-fashioned but I believe that
origin-server adaptation based on established specifications and
standards is the soundest and most widely available approach.  Now, this
is not to say that a stored set of declarations similar to the Document
Type Definition is not a useful approach, but who (or which entity) will
be responsible for the definition, maintenance and updating of such a
repository if the intent is to cater to every conceivable mobile device
available?

Be Well All,

Sotiris Sotiropoulos

Received on Sunday, 7 May 2006 20:35:54 UTC