RE: suggestions for new roles and properties in ARIA next

I support having aria-description but not as a means to provide 
descriptions of graphics that need long descriptions. I believe long 
descriptions should be in structured text, which cannot be provided in an 
attribute. In addition, attribute strings that are long can be 
problematic.

The primary use case for aria-description would be, in my mind, to provide 
the screen reader equivalent to title. Turning on title reading in your 
screen reader makes the web a horribly verbose experience, yet there are 
many instances where title text is directed specifically at AT or keyboard 
users. 
Authors could, in these cases, provide aria-description that is equal to 
title.

Also, there are situations where providing context awareness with some 
brief descriptive text directed at non-visual users can dramatically 
improve usability. Today, the only way to do this is to use 
aria-describedby to refer to text that is hidden off left, top, or right 
with css positioning -- a hack and a lot of extra work.

It would be up to the AT to determine how to render aria-description, but 
personally, I would want it treated in a manner similar to or identical to 
aria-describedby. JAWS, for example, automatically reads aria-describedby 
after the label, role, and value when tabbing to an element. It can be 
repeated with insert+tab. I would one day like to see it and all the other 
insert+tab info rendered in a virtual buffer with 2 quick presses of 
insert+tab.

Matt King
IBM I/T Chief Accessibility Strategist
IBM BT/CIO - Global Workforce and Web Process Enablement 
Phone: (503) 578-2329, Tie line: 731-7398
mattking@us.ibm.com



"John Foliot" <jfoliot@stanford.edu> 
Sent by: wai-xtech-request@w3.org
04/20/2011 01:35 PM

To
<sailesh.panchang@deque.com>, "'James Craig'" <jcraig@apple.com>, "'Steve 
Faulkner'" <faulkner.steve@gmail.com>
cc
"'W3C WAI-XTECH'" <wai-xtech@w3.org>, Richard 
Schwerdtfeger/Austin/IBM@IBMUS
Subject
RE: suggestions for new roles and properties in ARIA next






Sailesh Panchang wrote:
> 
> John,
> So is aria-description like a aria-label (or title attribute) ... i.e.
> it is not displayed on screen?

Hi Sailesh,

At this time aria-description is an idea only, so there are many good
questions still to be answered. I believe that as Steve Faulkner (and I)
are thinking, it would behave very similarly to @alt, in that under most
circumstances it would not display on screen.

> Then, is it available all the time like the alt to the screen reader
> user or
> can it be accessed only when one chooses to do so  like the longdesc?

I think that this would be at the discretion of the consuming tool -
screen readers could likely do either of those two options. My suspicion
is that the latter (behave like @longdesc) would be more user-friendly,
but I am not 100% sure. I'll turn the question around and ask you (as a
known daily screen reader user) - what behavior do you think *you'd*
prefer?

> Then
> one needs to be  aware that a particular img has a more detailed
> description.
> Or will it be rendered in the browser so that some sighted  mouse users
> or keyboard users can access it if they wish?

Again, this could/would be left to the discretion of the User Agents. As
is the case in some browsers today for longdesc, I personally think that
for sighted users access to the value string should be exposed via the
contextual menu, as a choice to consume or not consume. But that's just me
<grin>.

What are *your* thoughts?

JF


> Thanks,
> Sailesh Panchang
> Tel 571-449-3576
> 
> -----Original Message-----
> From: wai-xtech-request@w3.org [mailto:wai-xtech-request@w3.org] On
> Behalf
> Of John Foliot
> Sent: Wednesday, April 20, 2011 3:39 PM
> To: 'James Craig'; 'Steve Faulkner'
> Cc: 'W3C WAI-XTECH'; 'Richard Schwerdtfeger'
> Subject: RE: suggestions for new roles and properties in ARIA next
> 
> James Craig wrote:
> >
> > >> Kind of seems like you're stretching description to support the
> > semantics of @summary and maybe @longdesc. What makes you certain
> this
> > necessary?
> > >
> > > my thinking is that there are use cases where a developer would
> want
> > > to provide info to AT users which is redundant for non AT users,
> but
> > > is more a description than a name. having to place this information
> > > somewhere else and then refernece via describedby can be a pain.
> >
> > Can you give a specific example (using real text, not lorem ipsum) of
> > where this might be needed? I'd just like to come up with a
> defensible
> > case that cannot be done another way, or one that would gain
> > significant benefit over the alternative if done this way.
> 
> Hi James,
> 
> Taking a stab at this request:
> 
> <img
> src="http://macdailynews.files.wordpress.com/2011/02/110201_nielsen_01.
> png
> " alt="Line Graph - Mobile Operating System Share, US Q4 '09 - Q4 '10"
> aria-description="The graph shows the rapid growth of Android from 2%
> to
> 27% market share, a 11% decline for Blackberry OS to 27%, and a
> relatively
> flat pattern for iOS at 28% over a 14 month period.">
> 
> (image source:
> http://macdailynews.com/2011/02/01/nielsen_q410_u-s-
> _mobile_os_market_shar
> e_apple_ios_28_rim_blackberry_27_goo/)
> 
> I note that data supplied in the aria-description - notably the before
> and
> after percentages - are not spelled out in the associated article, as
> those numbers are 'obvious' to sighted users reading the article.
> 
> Would this meet your use-case request?
> 
> JF
> 
>

Received on Wednesday, 20 April 2011 22:15:50 UTC