RE: Question about restitution of aria-describedby

Take a look at the comment in 4.3.1 <https://www.w3.org/TR/accname-1.1/#terminology>  under point 2:

 

“By default, assistive technologies do not relay hidden information, but an author can explicitly override that and include hidden text as part of the accessible name or accessible description by using aria-labelledby or aria-describedby.”

 

And in this context ‘hidden’ means “that the element is not visible, perceivable, or interactive to any user.”

 

From: Armony ALTINIER - Koena <armony@koena.net> 
Sent: Tuesday, October 17, 2023 10:57 AM
To: w3c-wai-ig@w3.org
Subject: Re: Question about restitution of aria-describedby

 

Thanks a lot for your answers Marc and Nick.

But I am still confused. I read the Accessible Name and Description Computation, §4.2 <https://www.w3.org/TR/accname-1.1/#mapping_additional_nd_description>  and §4.3 <https://www.w3.org/TR/accname-1.1/#mapping_additional_nd_te> , and I still don't see anything about the case where the description is properly hidden with aria-describedby or aria-hidden=true.

And when I read the description of aria-hidden attribute in the WAI-ARIA 1.2 doc <https://www.w3.org/TR/wai-aria-1.2/#aria-hidden> , it is clear that the hidden text is not supposed to be exposed to an accessibility API.

Does anyone have a specific doc about this case? It still appears to me as a bad practice to link a description to a text with aria-hidden=true or display:none. But it seems it works, so maybe it is supposed to be working. It just doesn't make sense to me.

Armony

Le 17/10/2023 à 11:08, Marc Haunschild a écrit :

It also has a strong use case, that hidden text is not announced - so you can switch it by changing the display property value with JavaScript to control which (of several) texts should be read - for example depending on user input - to name just one example. Especially because we have CSS styles to effectively hide texts only visually. For me this is a very inconsistent approach and I really hate it.  

 

But it is, like it is and that's how we have to use it...

 

--
Mit freundlichen Grüßen

Marc Haunschild - he / him - #gernPerDu
Prüfstelle im BIK-BITV-Prüfverbund

Marc Haunschild Accessibility Consulting
Sonnenhof 32
53119 Bonn

Telefon: 0170 8 64 00 63
Web: https://Accessibility.Consulting https://mhis.de 
Email: Marc.Haunschild@Accessibility.Consulting <mailto:Marc.Haunschild@Accessibility.Consulting> 





Am 17.10.2023 um 10:43 schrieb Marc Haunschild (Accessibility Consulting)  <mailto:marc.haunschild@accessibility.consulting> <marc.haunschild@accessibility.consulting>:

 

Hi Armony, 

 

This behavior is typical for most (all?) screenreaders. I was also surprised when I heard that for the first time.

 

I can’t remember which, but I think one SR-browser combination does not read the hidden texts and it’s considered a bug.

 

You might want to check the computation of accname if you want to know more about the spec, if you’re not familiar with it. But as far as I remember it is not about hidden text. Maybe this is way sr behave the way they do:



 <https://www.w3.org/TR/accname-1.1/> Accessible Name and Description Computation 1.1

 <https://www.w3.org/TR/accname-1.1/> w3.org

 <https://www.w3.org/TR/accname-1.1/> <favicon.ico>

 

--

Mit freundlichen Grüßen 

 

Marc Haunschild

https://Accessibility.Consulting





Am 17.10.2023 um 10:36 schrieb Armony ALTINIER - Koena  <mailto:armony@koena.net> <armony@koena.net>:

 

Hi everyone,

I have a question about how aria-describedby is supposed to be rendered or not by screen readers.

I'm submitting this test file (see attachement) with 3 cases:

1. text with aria-describedby is linked to non-hidden text
2. text with aria-describedby is linked to text hidden with aria-hidden
3. text with aria-descrivbedby is linked with text hidden with display:none.


In case 1, it is normal for the text to be rendered. It seems strange to me that this also works on cases 2 and 3, whereas it works fine with NVDA.

Do you have any documentation about this behavior? Is it "normal"? If so, I'm having trouble understanding the logic. If it's not normal, and there are screen reader developers among you, I'd love a little explanation of this choice (I'll go fishing for info elsewhere if need be).

Many thanks in advance for any insight and documentation :)

Best Regards,
Armony

-- 


  <https://koena.net/app/uploads/2022/01/signature_armony.png> 
 <https://www.linkedin.com/in/armonyaltinier/>   <https://twitter.com/armonyaltinier> 

Armony Altinier 
Fondatrice et Présidente |   <https://koena.net/app/uploads/2022/01/signature_koena_logo.png> 

  _____  

Téléphone  : +33 (0)9 72 63 21  28 
Site web  :  <https://koena.net/> koena.net 
Twitter  :  <https://twitter.com/koenaFR> @koenaFR 


Découvrez l'ensemble des  <https://koena.net/?page_id=14196> adresses mails utiles chez Koena et les personnes associées. 

<demo-aria-describedby-EN.html>

 

-- 


  <https://koena.net/app/uploads/2022/01/signature_armony.png> 
 <https://www.linkedin.com/in/armonyaltinier/>   <https://twitter.com/armonyaltinier> 

Armony Altinier 
Fondatrice et Présidente |   <https://koena.net/app/uploads/2022/01/signature_koena_logo.png> 

  _____  

Téléphone  : +33 (0)9 72 63 21  28 
Site web  :  <https://koena.net> koena.net 
Twitter  :  <https://twitter.com/koenaFR> @koenaFR 


Découvrez l'ensemble des  <https://koena.net/?page_id=14196> adresses mails utiles chez Koena et les personnes associées.