RE: Comments on Device Description Repository Requirements 1.0

If you are offering a product such as a "wallpaper" image for a mobile
phone, then as part of the product offering the vendor may need to
determine unambiguously the precise screen size. This is one example
where an exact query would be needed. In general, however, it would be
sufficient to get a "fairly reliable" indication of the device
capabliity in order to have a good chance of deliverying a good quality
response.
 
While it may seem obvious to say that there should be a general
guideline on how best-effort should be determined, discussions with
vendors of proprietary solutions that incorporate such a solution shows
that the issue is not so clear. If one single technique is agreed, then
this would be the one used by the DDR. If there are more than one, with
none dominant, then perhaps the DDR will have a default technique, and
some means by which the user could indicate (in a query) if an
alternative technique should be used. This is all about the
design/implementation of the DDR, and therefore a topic to be considered
at the workshop.
 
As a practical example of a best-effort technique, please refer to the
fallback mechanism employed by WURFL.
 
---Rotan

________________________________

From: Lior Sion [mailto:LiorS@flashnetworks.com] 
Sent: 07 May 2006 08:31
To: Rotan Hanrahan; Lior Sion; public-ddwg@w3.org
Subject: RE: Comments on Device Description Repository Requirements 1.0



Rotan, 

Thanks for the quick response. 

Regarding your 2nd comment, I'm still failing to understand how will
this work with the best effort approach. If device A identifies itself
as A, which doesn't exist on the repository, there should be a general
guideline on what best effort query will return, don't you agree?

Whatever this method is, and I'm sure there will be a lot of discussion
on it - I'm guessing that in today's open market most queries would be
of this type (best effort) - and the response must at least be
understood by the clients of the DDR.

Lior 

-----Original Message----- 
From: Rotan Hanrahan [mailto:Rotan.Hanrahan@MobileAware.com] 
Sent: Thursday, May 04, 2006 3:52 PM 
To: Lior Sion; public-ddwg@w3.org 
Subject: RE: Comments on Device Description Repository Requirements 1.0 

There are two Requirements identified in the document: 

[DDR-QRY-EXACT] It SHOULD be possible for an Actor to query the DDR
specifying an exact match required in response to the query

[DDR-QRY-BEST-EFFORT] It SHOULD be possible for an Actor to query the
DDR specifying a best-effort match required in response to the query.

In the case of an exact match query, a failure to match would be
reported as a "miss", as you call it. In the case of a best-effort match
query, the response would be the nearest matching device. There are a
number of implementation possibilities regarding the representation of
approximation. The specific interpretation of "closest" would be
something we would consider in the implementation workshop. If more than
one interpretation is necessary, a means for the Actor to indicate the
required interpretation would be necessary. I do not wish to go into the
specifics of implementations at this point. Instead I wish just to
confirm that the issue you have mentioned is one that the group is aware
of, and wishes to consider as part of future implementation strategies.

Regarding the device/browser distinction, we acknowledge that
information about both may be necessary to achieve proper content
adaptation. We would only expect a device to be identified if the client

(browser) specfically indicates the device in which it is executing, or
if the delivery environment has some alternative means of detecting the
hardware during the interaction between client and server. The specifics
of the recognition process are out of scope for the DDR, which merely
assumes that a means of uniquely identifying the client/device exists,
and that the representation of the unique identity can be conveyed to
the DDR in queries.

Regards, 
---Rotan. 
________________________________ 

From: public-ddwg-request@w3.org [mailto:public-ddwg-request@w3.org] On
Behalf Of Lior Sion 
Sent: 04 May 2006 13:38 
To: public-ddwg@w3.org 
Subject: Comments on Device Description Repository Requirements 1.0 




Hi, 

I have a questions about v1.0 of the req: 

(a) 2.1.1, especially flow 2.1.7: what is the expected result of a
"miss" on the repository? Would the answer be "device not found" or will
the repository look for the closest know device? Will it mark the answer
as "closest"? How will this closest be found? 

I think that in today's market, and also in the foreseen future, this is
one of the main issues. Device identification changes, sometimes
slightly - different devices come with identical browsers - what would
count as a "hit"? The correct browsers version or the device version? 

Thanks, 

Lior Sion 



************************************************************************

************ 
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals &
computer viruses.

************************************************************************

************ 




  
************************************************************************
************ 
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals &
computer viruses.

************************************************************************
************ 



************************************************************************
************
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals &
computer viruses.
************************************************************************
************