W3C home > Mailing lists > Public > wai-xtech@w3.org > February 2012

Re: The 9 Citations Buttressing HTML's Issue-204

From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Date: Sat, 18 Feb 2012 02:38:24 +0100
To: Janina Sajka <janina@rednote.net>
Cc: "wai-xtech@w3.org" <wai-xtech@w3.org>
Message-ID: <20120218023824151139.594234f6@xn--mlform-iua.no>
Janina Sajka, Thu, 16 Feb 2012 23:07:42 -0500:

> and reached such a markedly different conclusion from that
> of the ARIA Task Force itself,

Conclusion about what? 
And where is the ARIA Task Force' conclusion w.r.t. ISSUE-204? In the 
ARIA specs, you mean?

I offer comments to the urls you pointed. See below.

> http://www.w3.org/WAI/PF/aria-implementation/#include_elements


This confirms that elements with @hidden must not be included in the 
a11y three. The HTMLwg is in violent *agreement* with ARIA on this, as 
much as I can see.

> 
http://www.w3.org/WAI/PF/aria-implementation/#mapping_additional_relations_reverse_relations

> 
http://www.w3.org/WAI/PF/aria-implementation/#mapping_additional_relations_reverse_relations


This seems quite important - so I quote: 

"Note that aria-describedby may reference structured or interactive 
information where users would want to be able to navigate to different 
sections of content. User agents MAY provide a way for the user to 
navigate to structured information referenced by aria-describedby and 
assistive technology SHOULD provide such a method."

This quote *does* say that AT/UA should make available to the user the 
interactive and structured semantics of an section that 
aria-describedby points to.
  
> http://www.w3.org/WAI/PF/aria-implementation/#mapping_role_table


No comment.

> http://www.w3.org/WAI/PF/aria-practices/#Descriptions_external


Quote: "if you wish to reference an external resource with 
aria-describedby, you can reference a link that in turn references the 
external resource"

And then it offers this this example, which I simplify for brevity:

   <img src=histogram
       alt="Histogram of Blackberry tree heights"
       aria-describedby=longdesc >
   <a id=longdesc href=link target=_description>Histogram data</a>


Comment: I have earlier proposed to change this text - even to remove 
it - because it did not speak to the facts. But now, I have heard from 
Jonas, that this technique already works, for visible elements. [Though 
when I asked, he did not point me to a particular build of Firefox for 
testing.] What ISSUE-204 promises is that what the ARIA Practises 
document here describe for a link without HTML5's @hidden attribute, 
would also be possible for a link with HTML5's @hidden attribute.

BUT NOTE: The link inside this example could very well be hidden, 
despite the fact that does not have aria-hidden=true. For instance: It 
might have been placed off-screen via CSS. The example does not say 
anything about this. [But note that ARIA 1.0 says that when an element 
is hidden, then it should have aria-hidden=true. So it is quite 
possible that this element should have had that. Because, as long as it 
is not hidden, it would be read read twice by the AT: Once when 
presenting the aria-describedby relationship and once when the 
rendering proceeds to the next element,after the <img>.]
 
> http://www.w3.org/WAI/PF/aria-practices/#Descriptions_tooltip


Hm. I guess a role=tooltip element could contain a link to a long 
description ...

> http://www.w3.org/WAI/PF/aria-practices/#kbd_layout_remaining_description

> http://www.w3.org/WAI/PF/aria-practices/#kbd_layout_remaining_description


Quote: "This is unlike longdesc which typically requires the author to 
create a separate file to describe a picture." 

Comment: Longdesc does not 'typically require this' - it is just that 
longdesc typically is *used* that way.

Quote: "It is preferable to have the descriptive text in prose as well 
so that it is readily available to all users"

Comment: Question: Why is it preferable, when this could lead to 
repetition? At any rate: That it is preferable, means that the text 
take account for the fact that some will hide the description.

Quote: "This is the preferred vehicle for providing long descriptions 
for elements in your document. … snip … aria-describedby can also be 
used to point to a link to another page"

Comment: The sections discusses @longdesc many times. But says that 
@aria-describedby is the preferred method for long descriptions inside 
the document. And, despite that it claims that @longdesc's primary role 
is to point to other documents, it does demonstrate how to do the same 
with @aria-describedby. 

This document several times demonstrates how to use the alt attribute 
correctly. BUT NOT A SINGLE TIME DOES IT DEMONSTRATE HOW TO USE 
@longdesc - not even when it speaks about pointing to external 
documents.

The elephant in the room .... No! The elephants - in plural. If the 
ARIA Task Force has reached a 'different conclusion' about - hm - 
@longdesc, then it has done its best to hide it.

> http://www.w3.org/WAI/PF/aria/states_and_properties#aria-describedby


No comment. But I would like to also point to ARIA 1.0's section on 
Text Alternative Computation:

http://www.w3.org/WAI/PF/aria/roles#textalternativecomputation


quote: "Skip hidden elements unless the author specifies to use them 
via an aria-labelledby or aria-describedby being used in the current 
computation."
-- 
Leif Halvard Silli
Received on Saturday, 18 February 2012 01:38:56 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:25:31 UTC