Re: Split Issue 30? (Now: ISSUE-204)

Ian Hickson Wed, 15 Feb 2012 02:13:14 +0000 (UTC):
On Mon, 13 Feb 2012, Maciej Stachowiak wrote:
>> I think this is what Jonas is proposing - behavior for @hidden elements 
>> that would not apply to content which is set to display: none in other 
>> ways. I don't think be advocates changing behavior for display: none 
>> content in any way.
> I don't think it would make sense for content that is marked up as 
> relevant (not hidden="") but happens to be display:none to be less 
> accessible than content that is marked as irrelevant (hidden="") and thus 
> display:none, FWIW.


If I have gotten this right: Aria-describedby=hiddenSection would be 
presented as if it was a visible section. While 
aria-describedby=displayNoneSection would get a text string 
presentation. Thus, it seems like adding @hidden, could make 
non-displayed sections more accessible or cause a more "semantic" 

But w.r.t. what you said about "continuations"/"a block is nested 
inside an inline" [1], then does the fallback of <object data=image.gif 
> count as a "shadow DOM" too? And did you work on fixing fallback of 
<object> as well … ? After all, <object> defaults to display:inline and 
it its content is hidden (but not irrelevant). (My question relates to 
ISSUE-158 [2].)

It seems contrived if AT would treat 
   <object aria-describedby=x >
      <div id=x hidden><p>Foo
different from how
   <object aria-describedby=x >
      <div id=x hidden><p>Foo
is treated. If fallback of <object> gets treated like the <canvas> 
shadow DOM, then adding @hidden should not gain any thing. Which in 
turn would diminish the risk that authors perceive it as if @hidden can 
make invisible elements more "accessible". A question that pops up: Are 
there elements - such as children of <object> - where @hidden does not 
make sense?



Leif Halvard Silli

Received on Wednesday, 15 February 2012 04:56:21 UTC