E-mail List Archives
Thread: Is ARIA required to conform to WCAG?
Number of posts in this thread: 12 (In chronological order)
From: Karen McCall
Date: Sun, Mar 26 2023 8:53AM
Subject: Is ARIA required to conform to WCAG?
No previous message | Next message →
Morning Everyone!
1. Is the use of ARIA attributes mandatory to conform to WCAG 2.x or are ARIA attributes optional? I can't find a definitive answer.
* Can I have a WCAG 2.x conforming web site without using ARIA?
2. If an ARIA attribute is used, does it affect the type of HTML tag used? For example, if a you have an <H1> tag and use ARIA attributes, does the tag now become a <P> Tag because of the ARIA attributes?
Cheers, Karen
From: Dean.Vasile
Date: Sun, Mar 26 2023 8:54AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
No aria is not required.
In fact, if you can do it without using aria, it is preferred.
Dino
617-799-1162
Dino's Canteen 1618
11 Eglin St,
Hanscom AFB
Bedford, MA
> On Mar 26, 2023, at 10:53 AM, Karen McCall < = EMAIL ADDRESS REMOVED = > wrote:
>
> Morning Everyone!
>
> 1. Is the use of ARIA attributes mandatory to conform to WCAG 2.x or are ARIA attributes optional? I can't find a definitive answer.
> * Can I have a WCAG 2.x conforming web site without using ARIA?
> 2. If an ARIA attribute is used, does it affect the type of HTML tag used? For example, if a you have an <H1> tag and use ARIA attributes, does the tag now become a <P> Tag because of the ARIA attributes?
> Cheers, Karen
> > > >
From: Patrick H. Lauke
Date: Sun, Mar 26 2023 9:33AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
On 26/03/2023 15:53, Karen McCall wrote:
> Morning Everyone!
>
> 1. Is the use of ARIA attributes mandatory to conform to WCAG 2.x or are ARIA attributes optional? I can't find a definitive answer.
They're required *if* what you're building needs to expose its
name/role/state/value, but doesn't by default. Like a button that acts
as a toggle needs to expose when it's pressed or not pressed, so unless
you add aria-pressed="true"/aria-pressed="false" to expose its state
programmatically, you're failing 4.1.2 at least.
> * Can I have a WCAG 2.x conforming web site without using ARIA?
You can if you don't use/build any custom controls that make no
programmatic sense without ARIA.
> 2. If an ARIA attribute is used, does it affect the type of HTML tag used? For example, if a you have an <H1> tag and use ARIA attributes, does the tag now become a <P> Tag because of the ARIA attributes?
Depends which attribute. if you're adding a role, like <h1
role="button"> for instance, you've just told the browser to expose that
heading as a button. If you don't force a role, no, your elements will
remain the same, so you *can* add aria-* attributes without changing the
nature of the element (but of course, you need to actually know what
you're doing...can't just add aria attributes at random, they need to
still make sense...for instance, it would be utterly illogical - and
will likely be ignored by assistive technologies - if you did something
like <h1 aria-pressed="true"> because a heading can't have a state of
being pressed/not pressed.
I suggest learning more about aria itself and when it makes sense to use
it or not...
P
--
Patrick H. Lauke
https://www.splintered.co.uk/ | https://github.com/patrickhlauke
https://flickr.com/photos/redux/ | https://www.deviantart.com/redux
twitter: @patrick_h_lauke | skype: patrick_h_lauke
From: Birkir R. Gunnarsson
Date: Sun, Mar 26 2023 9:43AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
If you can use the right HTML for all the content, structure and
functionality you need on the site, then no ARIA is required.
If you cannot use the right HTML markup, you can fix it with ARIA.
E.g.
If, for some reason, you have a button but it is coded with a <div>
element (that, in itself has no button role or function) you can
either replace it by a <button> element (which has all the function ou
need) ot turn it into a button using tabindex, JavaScript and ARIA
* add role="button" to make it announce as a button by a screen reader
* add tabindex="0" to make it focusable with keyboard
* Add JavaScript functions to let users activate the button with
spacebar or the enter key
Obviously this is a lot more work than just replacing the <div>
element with a <button> element.
There are a few things that ARIA can do that HTML cannot.
* Connecting form inputs and error messages (you need aria-escribedby
or aria-errormessage, HTML does not offer a way other than stuffing he
error message inside the <label>)
* Have screen readers announce an update automatically when it happens
(usign aria-live, role="alert" or role="status"). The only HTML to
achieve this is the <output> element and it does not work for all
situations
Build tabs and menus (HTML has no way to do thise).
You used to need ARIA to build accordions but now you can do it with
HTML using the <details> and <summary> element structure.
You used to require ARIA to biuld modal dialogs but finally, after 10
yars in the HTML spec, the support for the <dialog> element is
sufficient to use it.
You used to require ARIA to build a search input with suggestions that
are displayed as you type. Now you can do it with an <input> element,
a list attribute and a <datalist> element (support in most browsers,
though not perfect).
So, again, it depends what you are building on the website. You can do
90% of what you need without ARIA, but you may need it for forms,
tabs, menus and a few other things.
You may also need it if you cannot use the HTML you need (sometimes
you are building a website from frameworks that didn't build HTML
right so you need ARIA to correct it).
As for your second qustions. Yes, when you use ARIA it always
overrides the HTML you use, that's its purpose.
<h1 role="button">This is a heading but a screen reader will announce
it as a button</h1>
or even
<h1 role="heading" aria-level="5">This is coded as an h1 heading but
announced as an h5 heading by a screen reader</h1>
On 3/26/23, Patrick H. Lauke < = EMAIL ADDRESS REMOVED = > wrote:
>
>
> On 26/03/2023 15:53, Karen McCall wrote:
>> Morning Everyone!
>>
>> 1. Is the use of ARIA attributes mandatory to conform to WCAG 2.x or
>> are ARIA attributes optional? I can't find a definitive answer.
>
> They're required *if* what you're building needs to expose its
> name/role/state/value, but doesn't by default. Like a button that acts
> as a toggle needs to expose when it's pressed or not pressed, so unless
> you add aria-pressed="true"/aria-pressed="false" to expose its state
> programmatically, you're failing 4.1.2 at least.
>
>> * Can I have a WCAG 2.x conforming web site without using ARIA?
>
> You can if you don't use/build any custom controls that make no
> programmatic sense without ARIA.
>
>> 2. If an ARIA attribute is used, does it affect the type of HTML tag
>> used? For example, if a you have an <H1> tag and use ARIA attributes, does
>> the tag now become a <P> Tag because of the ARIA attributes?
>
> Depends which attribute. if you're adding a role, like <h1
> role="button"> for instance, you've just told the browser to expose that
> heading as a button. If you don't force a role, no, your elements will
> remain the same, so you *can* add aria-* attributes without changing the
> nature of the element (but of course, you need to actually know what
> you're doing...can't just add aria attributes at random, they need to
> still make sense...for instance, it would be utterly illogical - and
> will likely be ignored by assistive technologies - if you did something
> like <h1 aria-pressed="true"> because a heading can't have a state of
> being pressed/not pressed.
>
> I suggest learning more about aria itself and when it makes sense to use
> it or not...
>
> P
> --
> Patrick H. Lauke
>
> https://www.splintered.co.uk/ | https://github.com/patrickhlauke
> https://flickr.com/photos/redux/ | https://www.deviantart.com/redux
> twitter: @patrick_h_lauke | skype: patrick_h_lauke
> > > > >
--
Work hard. Have fun. Make history.
From: Karen McCall
Date: Sun, Mar 26 2023 9:54AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
Thank you all!
This is why I value participating in this list. I get great answers and examples for questions I ask!
Cheers, Karen
From: L Snider
Date: Sun, Mar 26 2023 10:51AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
I just saw a webpage that used aria labels on links, that in my view made
the accessible simple links way, way less accessible. ARIA landmarks are
used by screen reader users I know, and anything dynamic or changing, or
links that can't be given text. I see the overuse of ARIA a lot, especially
by those who didn't design in the HTML only days! Just my experience with
these ones...
On Sun, Mar 26, 2023 at 12:54 PM Karen McCall < = EMAIL ADDRESS REMOVED = > wrote:
> Thank you all!
>
> This is why I value participating in this list. I get great answers and
> examples for questions I ask!
>
> Cheers, Karen
>
>
From: Birkir R. Gunnarsson
Date: Sun, Mar 26 2023 11:02AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
aria-label is useful on links or buttons that are visually labeled by
icons (e.g. links to social media or buttons labeled with symbols,
like the search button or close button). aria-label is an easy way to
provide a screen reader / assistive technology label for those.
When a link already has sufficien text aria-label is redundant. Even
if the text is insufficient there are other ways, e.g. adding visually
hidden text, to achieve the purpose without aria. You could use
aria-label but then you have to take great care that the aria-label
value includes the visible text.
e.g.
<a href="http://www.google." aria-label="Google">Click here</a> is
invalid use of aria-label, but
<a href="http://www.google.com" aria-label="Click ehre to visit
Google">Click here</a> is acceptable.
On 3/26/23, L Snider < = EMAIL ADDRESS REMOVED = > wrote:
> I just saw a webpage that used aria labels on links, that in my view made
> the accessible simple links way, way less accessible. ARIA landmarks are
> used by screen reader users I know, and anything dynamic or changing, or
> links that can't be given text. I see the overuse of ARIA a lot, especially
> by those who didn't design in the HTML only days! Just my experience with
> these ones...
>
> On Sun, Mar 26, 2023 at 12:54 PM Karen McCall < = EMAIL ADDRESS REMOVED = > wrote:
>
>> Thank you all!
>>
>> This is why I value participating in this list. I get great answers and
>> examples for questions I ask!
>>
>> Cheers, Karen
>>
>>
From: Geethavani.Shamanna
Date: Mon, Mar 27 2023 4:58AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
On a related note: My question is more about the terminology.
Although the words 'required' and 'normative' do not convey the same meaning, they are sometimes used interchangeably. For example, are WCAG sufficient techniques 'required', as they are not normative?
Geetha
From: Mark Magennis
Date: Mon, Mar 27 2023 5:27AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
It isn't required to use any sufficient technique. Although using one would be considered enough to pass the SC you can use any technique of your own as long as it works.
From: Patrick H. Lauke
Date: Mon, Mar 27 2023 5:31AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
On 27/03/2023 11:58, Geethavani.Shamanna wrote:
> On a related note: My question is more about the terminology.
> Although the words 'required' and 'normative' do not convey the same
meaning, they are sometimes used interchangeably. For example, are WCAG
sufficient techniques 'required', as they are not normative?
Techniques are informative, non-normative. And they're far from
comprehensive. They show just a handful of ways in which an SC can be
passed or failed (and many are in dire need of either being retired or
updated). What counts is how you satisfy the normative ask of any
particular SC ... even if you use a method/way that is not even
mentioned at all in the understanding/techniques, as long as you are
confident/can demonstrate that it satisfies the normative requirements
of the SC itself, you're good.
P
--
Patrick H. Lauke
https://www.splintered.co.uk/ | https://github.com/patrickhlauke
https://flickr.com/photos/redux/ | https://www.deviantart.com/redux
twitter: @patrick_h_lauke | skype: patrick_h_lauke
From: Birkir R. Gunnarsson
Date: Mon, Mar 27 2023 9:41AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | Next message →
This is mostly true, though a little bit confusing.
Sufficient techniques are normative in that they are a documented way
to pass a WCAG success criterion.
They're not the only way to pass one, but if you follow that technique
exactly, WCAG is saying that you pass.
Similarly, a WCAG failing technique is normative in the sense that it
is documented as failing one or more WCAG success criteria. It's not
the only way to fail one, oh no precious.
Given how infrequently those techniques are updated I've always found
this a bit troublesome.
On 3/27/23, Patrick H. Lauke < = EMAIL ADDRESS REMOVED = > wrote:
> On 27/03/2023 11:58, Geethavani.Shamanna wrote:
> > On a related note: My question is more about the terminology.
> > Although the words 'required' and 'normative' do not convey the same
> meaning, they are sometimes used interchangeably. For example, are WCAG
> sufficient techniques 'required', as they are not normative?
>
> Techniques are informative, non-normative. And they're far from
> comprehensive. They show just a handful of ways in which an SC can be
> passed or failed (and many are in dire need of either being retired or
> updated). What counts is how you satisfy the normative ask of any
> particular SC ... even if you use a method/way that is not even
> mentioned at all in the understanding/techniques, as long as you are
> confident/can demonstrate that it satisfies the normative requirements
> of the SC itself, you're good.
>
> P
> --
> Patrick H. Lauke
>
> https://www.splintered.co.uk/ | https://github.com/patrickhlauke
> https://flickr.com/photos/redux/ | https://www.deviantart.com/redux
> twitter: @patrick_h_lauke | skype: patrick_h_lauke
> > > > >
--
Work hard. Have fun. Make history.
From: Patrick H. Lauke
Date: Mon, Mar 27 2023 9:55AM
Subject: Re: Is ARIA required to conform to WCAG?
← Previous message | No next message
On 27/03/2023 16:41, Birkir R. Gunnarsson wrote:
> This is mostly true, though a little bit confusing.
>
> Sufficient techniques are normative in that they are a documented way
> to pass a WCAG success criterion.
> They're not the only way to pass one, but if you follow that technique
> exactly, WCAG is saying that you pass.
>
> Similarly, a WCAG failing technique is normative in the sense that it
> is documented as failing one or more WCAG success criteria. It's not
> the only way to fail one, oh no precious.
Sure, but I'd be careful with "normative in the sense that..." because
that then is at odds with the wording of WCAG itself, which makes the
distinction between the *normative* part only being the core WCAG
document itself
https://www.w3.org/TR/WCAG21/#interpreting-normative-requirements, and
both the Understanding https://www.w3.org/WAI/WCAG21/Understanding/ and
Techniques
https://www.w3.org/WAI/WCAG21/Understanding/understanding-techniques#techniques-are-informative
are signposted as *not* being normative / being informative.
By reimagining what "normative" actually means, this might actually make
matters more confusing.
P
--
Patrick H. Lauke
https://www.splintered.co.uk/ | https://github.com/patrickhlauke
https://flickr.com/photos/redux/ | https://www.deviantart.com/redux
twitter: @patrick_h_lauke | skype: patrick_h_lauke