RE: Clarification of WCAG intent and meaning of techniques [Re: WCAG considering amending F65 to NOT fail missing ALT text if title or aria-label is present]

Schnabel, Stefan, Thu, 28 Nov 2013 12:48:49 +0000:
>>> So, to replace @alt with an @aria-* attribute, would be to do the 
> opposite of what the WCAG Robustness principle requires
> 
> No.

I maintain my general point that *relying* on @alt is more robust than 
*relying* on @aria-whatever. In your example, however, you don't rely 
on aria-labelledby. (Though you do rely on it for making a direct 
association between label and image.)

> <img src="../images/giraffe.jpg" aria-labelledby="123"/>
> <p id="123">Giraffe grazing on tree branches</p>
> 
> is equivalent since even if the image is missing the text describing 
> the image is still there. 

I came with a general claim about ”an @aria-* attribute”. (A ”perfect” 
example would be an <img> where one had replaced alt="Lorem ipsum" with 
aria-label="Lorem ipsum".) Whereas you came up with specific claim 
about *aria-labelledby*, where the the attribute points to a adjacent 
element.

You are right that in your particular example, users get to read the 
#123 paragraph even if image-display is disabled.

It would be nice if using aria-labelledby the way you do here, would 
also make AT *not* read the description twice. But if that is the 
intention, then the ARIA spec should specify that this is what is 
supposed to happen. It is not very robust to *hope* that AT don't 
repeat the text.
-- 
leif halvard silli