W3C home > Mailing lists > Public > w3c-wai-ua@w3.org > April to June 2014

Action-978 - response to MS05

From: Jeanne Spellman <jeanne@w3.org>
Date: Tue, 03 Jun 2014 12:04:12 -0400
Message-ID: <538DF1FC.604@w3.org>
To: User Agent Working Group <w3c-wai-ua@w3.org>
I found that I had written the response in the spreadsheet, but had not 
sent it to the group.


Response to MS05: UAWG  wrote the examples in Implementing UAAG to be 
illustrative of common uses by people with disabilities. In order to 
avoid being excessively prescriptive to the browser, we did not make 
divisions between platform, OS, browser and assistive technologies, 
knowing that the field is evolving rapidly, and the accessibility 
feature that is on the browser on one device, may be on the OS on 
another.  See the UAAG Conformance Applicability Note #7: "Relationship 
with operating system or platform: The user agent does not need to 
implement every behavior itself. A required behavior may be provided by 
the platform, user agent, user agent extensions, or potentially other 
layers. All are acceptable, as long as they are enumerated in the 
conformance claim."

The UAAG Conformance Applicability Note #1: "Recognized Content Only: 
UAAG 2.0 success criteria only apply to web content and its behaviors 
that can be recognized by user agents." UAWG has clarified the 
definition of "recognize".


MS05: Examples in the implementation document do not make distinction of 
content, browser, assistive technologies, and OS
We recognize the value of personalizing the examples, but these examples 
are not implementation examples. They are use case scenarios with no 
explanation of what is expected of the browser as oppose to that of the 
content, the OS, and the AT. It is understandable that average users do 
not understand the roles and responsibilities of content, browsers, 
assistive technologies, and OS. But the working group needs to provide 
more sufficient context for a technical audience. We believe this 
problem is a reflection of the lack of focus in UAAG. Take, for example, 
the second example from 1.2.2. (Maria uses a screen reader. When a table 
lacks marked up header rows, the user agent gives her the option to have 
the first row treated as the table header row.) It is entirely unclear 
as to what is expected to be done from the example. It is possible for 
the reader to interpret that: a) the browser is supposed to alter the 
content without authorization from the content author to change the 
structure of the table, or b) the browser is supposed to alter what it 
passes through to the accessibility API, or c) the AT is supposed to 
figure out the table header row from the accessibility API and present 
it to the user as such, or d) the AT is supposed to interrogate the HTML 
code, determine the header, ignore the information from an accessibility 
API, and present the header to the user Since there are so many ways to 
interpret the example and the lack of scope of what a user agent is, it 
is essentially not implementable in reality. Microsoft asks the working 
group to reexamine the fundamental nature of what a user agent is and 
rewrite UAAG from top to bottom with a more precise understanding of 
user agents and what can be done to make them more accessible.
Received on Tuesday, 3 June 2014 16:03:59 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:49:45 UTC