WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: hreflang underestimated?

for

Number of posts in this thread: 6 (In chronological order)

From: wolfgang.berndorfer@zweiterblick.at
Date: Mon, Feb 21 2022 8:01AM
Subject: hreflang underestimated?
No previous message | Next message →

SR seem to ignore the html attribute *hreflang*. (Tested JAWS and NVDA)



hreflang gives a hint to the language of the resource in a link.

https://html.spec.whatwg.org/#attr-hyperlink-hreflang



IMO this hint would be small and fine for SR too:

* Get immediate info, not frustration after clicking. (--> link
purpose)
* Get globally the same info, not depending on work arounds.



Vendors feedback was: Too noisy and not for AT but for SEO .



What do UAAG require from AT? Support each html attribute?

What do you think about the relevance for users?



Wolfgang

From: Patrick H. Lauke
Date: Mon, Feb 21 2022 8:53AM
Subject: Re: hreflang underestimated?
← Previous message | Next message →

On 21/02/2022 15:01, = EMAIL ADDRESS REMOVED = wrote:
> SR seem to ignore the html attribute *hreflang*. (Tested JAWS and NVDA)
>
>
>
> hreflang gives a hint to the language of the resource in a link.
>
> https://html.spec.whatwg.org/#attr-hyperlink-hreflang
>
>
>
> IMO this hint would be small and fine for SR too:
>
> * Get immediate info, not frustration after clicking. (--> link
> purpose)
> * Get globally the same info, not depending on work arounds.

Note that it would be odd in this case to have SR users get more useful
info than non-SR users, as this could potentially benefit all users in
general, users with cognitive disabilities, etc.

> What do UAAG require from AT? Support each html attribute?

No, not really. See https://www.w3.org/TR/UAAG20/

Also, anecdotally, browser/AT companies don't really care much about
UAAG (and as there's no legal or legal-adjacent mandate for them to
follow it per se, the same way that WCAG has been included/referenced by
various pieces of legislation, they have little incentive to)

> What do you think about the relevance for users?

I think it would be a nice little piece of information for the browser
to surface to all users (not just SR users). But its real-world usage is
so low that I doubt they'll invest much time in implementing anything.
Chicken and egg. Might be relegated to something like a browser extension.

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: wolfgang.berndorfer@zweiterblick.at
Date: Mon, Feb 21 2022 9:43AM
Subject: Re: hreflang underestimated?
← Previous message | Next message →

Hi Patrick,
I stumbled over the attribute in a .css, which was something like:
a[hreflang="en"] {
background-image: url('english_icon.png');
padding-left: 30px;
…
}

Then I tested with SR and wondered, why the attribute was ignored.
So, the oddness was the other way round.

A visual hint might be a choice for browsers.
But an interpretation for SR-users is a duty of the SR. “Link *German* Oh wie schön ist Panama”
Compare *required*: It is announced by SR but needs a work around for sighted people.
That's no argument against SR interpretation.

Wolfgang

-----Original Message-----
From: WebAIM-Forum < = EMAIL ADDRESS REMOVED = > On Behalf Of Patrick H. Lauke
Sent: Monday, February 21, 2022 4:53 PM
To: = EMAIL ADDRESS REMOVED =
Subject: Re: [WebAIM] hreflang underestimated?

On 21/02/2022 15:01, = EMAIL ADDRESS REMOVED = wrote:
> SR seem to ignore the html attribute *hreflang*. (Tested JAWS and
> NVDA)
>
>
>
> hreflang gives a hint to the language of the resource in a link.
>
> https://html.spec.whatwg.org/#attr-hyperlink-hreflang
>
>
>
> IMO this hint would be small and fine for SR too:
>
> * Get immediate info, not frustration after clicking. (--> link
> purpose)
> * Get globally the same info, not depending on work arounds.

Note that it would be odd in this case to have SR users get more useful info than non-SR users, as this could potentially benefit all users in general, users with cognitive disabilities, etc.

> What do UAAG require from AT? Support each html attribute?

No, not really. See https://www.w3.org/TR/UAAG20/

Also, anecdotally, browser/AT companies don't really care much about UAAG (and as there's no legal or legal-adjacent mandate for them to follow it per se, the same way that WCAG has been included/referenced by various pieces of legislation, they have little incentive to)

> What do you think about the relevance for users?

I think it would be a nice little piece of information for the browser to surface to all users (not just SR users). But its real-world usage is so low that I doubt they'll invest much time in implementing anything.
Chicken and egg. Might be relegated to something like a browser extension.

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: Patrick H. Lauke
Date: Mon, Feb 21 2022 9:56AM
Subject: Re: hreflang underestimated?
← Previous message | Next message →

On 21/02/2022 16:43, = EMAIL ADDRESS REMOVED = wrote:
> Hi Patrick,
> I stumbled over the attribute in a .css, which was something like:
> a[hreflang="en"] {
> background-image: url('english_icon.png');
> padding-left: 30px;
> …
> }
>
> Then I tested with SR and wondered, why the attribute was ignored.
> So, the oddness was the other way round.

So based on the above, a failure of 1.1.1 on the part of the author.

> A visual hint might be a choice for browsers.
> But an interpretation for SR-users is a duty of the SR. “Link *German* Oh wie schön ist Panama”
> Compare *required*: It is announced by SR but needs a work around for sighted people.
> That's no argument against SR interpretation.

*shrug* well, if you can convince SR companies that there's value in
supporting/announcing this...cool. Probably as an option/setting. But
again, chicken and egg, as there's not much use out there in the wild
web for that attribute...

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: wolfgang.berndorfer@zweiterblick.at
Date: Tue, Feb 22 2022 2:11AM
Subject: Re: hreflang underestimated?
← Previous message | Next message →

There are a lot of use cases for hreflang. Let us think about multi-language countries like Switzerland or Canada. And the whole www is multilanguage happily. Perhaps the English-speaking community sees less need for the attribute, but there are the developers.

1.1.1: What should we tell an author, who states that the background image would have a text alternative, if only SR would do their job? Frustrating!

Perhaps really the browsers should give a hint. We need a fallback mechanism anyway when Stylesheets are disabled.


-----Original Message-----
From: WebAIM-Forum < = EMAIL ADDRESS REMOVED = > On Behalf Of Patrick H. Lauke
Sent: Monday, February 21, 2022 5:56 PM
To: = EMAIL ADDRESS REMOVED =
Subject: Re: [WebAIM] hreflang underestimated?

On 21/02/2022 16:43, = EMAIL ADDRESS REMOVED = wrote:
> Hi Patrick,
> I stumbled over the attribute in a .css, which was something like:
> a[hreflang="en"] {
> background-image: url('english_icon.png');
> padding-left: 30px;
> …
> }
>
> Then I tested with SR and wondered, why the attribute was ignored.
> So, the oddness was the other way round.

So based on the above, a failure of 1.1.1 on the part of the author.

> A visual hint might be a choice for browsers.
> But an interpretation for SR-users is a duty of the SR. “Link *German* Oh wie schön ist Panama”
> Compare *required*: It is announced by SR but needs a work around for sighted people.
> That's no argument against SR interpretation.

*shrug* well, if you can convince SR companies that there's value in supporting/announcing this...cool. Probably as an option/setting. But again, chicken and egg, as there's not much use out there in the wild web for that attribute...

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: Patrick H. Lauke
Date: Tue, Feb 22 2022 3:11AM
Subject: Re: hreflang underestimated?
← Previous message | No next message

On 22/02/2022 09:11, = EMAIL ADDRESS REMOVED = wrote:
> There are a lot of use cases for hreflang. Let us think about multi-language countries like Switzerland or Canada. And the whole www is multilanguage happily. Perhaps the English-speaking community sees less need for the attribute, but there are the developers.

You don't need to convince *me* of it ;)

But as said, it's a chicken and egg problem, as is often the case.
Developers will tend not to use it because there's no sensible
browser/AT support, and browsers/ATs won't invest time in supporting it
as long as it's so niche and used so infrequently.

> 1.1.1: What should we tell an author, who states that the background image would have a text alternative, if only SR would do their job? Frustrating!

That hreflang in itself is not accessibility supported, sadly, as
neither the browser nor AT do anything with it / expose it in any way at
the moment. Sadly, but that's the reality on the ground.

--
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