WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: How useful is the datalist element?

for

Number of posts in this thread: 5 (In chronological order)

From: Jeremy Echols
Date: Fri, May 26 2017 4:20PM
Subject: How useful is the datalist element?
No previous message | Next message →

It seems like being native HTML, it would be pretty accessible, but in my testing with Firefox and NVDA, it seems like it's... just painful to use. For instance, NVDA doesn't notify me when typeahead options show up, but worse, when I use the down-arrow to start selecting items, it re-reads the input label and aria-describedby text... on every single down-arrow selection.

Are there JavaScript typeaheads which are better? Is support for datalist just at such a bad point that it shouldn't be used?

From: Chaals is Charles McCathie Nevile
Date: Fri, May 26 2017 4:31PM
Subject: Re: How useful is the datalist element?
← Previous message | Next message →

Interesting question - coincidentally, today someone proposed a
different improvement by allowing grouping:
https://github.com/w3c/html/issues/932

My sense is that implementation is still immature. And not very good.
File bugs on browsers...

cheers

On 27/05/17 00:20, Jeremy Echols wrote:
> It seems like being native HTML, it would be pretty accessible, but in my testing with Firefox and NVDA, it seems like it's... just painful to use. For instance, NVDA doesn't notify me when typeahead options show up, but worse, when I use the down-arrow to start selecting items, it re-reads the input label and aria-describedby text... on every single down-arrow selection.
>
> Are there JavaScript typeaheads which are better? Is support for datalist just at such a bad point that it shouldn't be used?
> > > > --
Charles McCathie Nevile - standards - Yandex
= EMAIL ADDRESS REMOVED = - Find more at http://yandex.com

From: Jeremy Echols
Date: Fri, May 26 2017 4:43PM
Subject: Re: How useful is the datalist element?
← Previous message | Next message →

This is very disappointing to hear. Thanks for the response.

I don't know enough about the accessibility APIs to know what kind of bug to file. I'm assuming the browser is sending the wrong information or too much information or something, but is there some way to see exactly what NVDA is being told by the browser? And I'm not sure what the expected behavior would even be. It seems that the concept of a "combo box" hasn't been filled in yet by the w3c "aria practices" authors: https://github.com/w3c/aria-practices/issues/31.

-----Original Message-----
From: WebAIM-Forum [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Chaals is Charles McCathie Nevile
Sent: Friday, May 26, 2017 3:32 PM
To: WebAIM Discussion List < = EMAIL ADDRESS REMOVED = >
Subject: Re: [WebAIM] How useful is the datalist element?

Interesting question - coincidentally, today someone proposed a different improvement by allowing grouping:
https://github.com/w3c/html/issues/932

My sense is that implementation is still immature. And not very good.
File bugs on browsers...

cheers

On 27/05/17 00:20, Jeremy Echols wrote:
> It seems like being native HTML, it would be pretty accessible, but in my testing with Firefox and NVDA, it seems like it's... just painful to use. For instance, NVDA doesn't notify me when typeahead options show up, but worse, when I use the down-arrow to start selecting items, it re-reads the input label and aria-describedby text... on every single down-arrow selection.
>
> Are there JavaScript typeaheads which are better? Is support for datalist just at such a bad point that it shouldn't be used?
> > > archives at http://webaim.org/discussion/archives
> --
Charles McCathie Nevile - standards - Yandex
= EMAIL ADDRESS REMOVED = - Find more at http://yandex.com

From: Birkir R. Gunnarsson
Date: Sun, May 28 2017 10:31AM
Subject: Re: How useful is the datalist element?
← Previous message | Next message →

How are you using the <datalist> element?
Are you using it with the list attribute of the <input> element?


I think it is going to be highly useful but, just like with many HTML5
elements and attributes, the browser and assistive technology support
is not quite there yet.
It is a frustrating journey, but if we can keep filing bugs and have
this fixed, the payoff in terms of accessibility of complex components
is going to be enormous.


On 5/26/17, Jeremy Echols < = EMAIL ADDRESS REMOVED = > wrote:
> This is very disappointing to hear. Thanks for the response.
>
> I don't know enough about the accessibility APIs to know what kind of bug to
> file. I'm assuming the browser is sending the wrong information or too much
> information or something, but is there some way to see exactly what NVDA is
> being told by the browser? And I'm not sure what the expected behavior
> would even be. It seems that the concept of a "combo box" hasn't been
> filled in yet by the w3c "aria practices" authors:
> https://github.com/w3c/aria-practices/issues/31.
>
> -----Original Message-----
> From: WebAIM-Forum [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf
> Of Chaals is Charles McCathie Nevile
> Sent: Friday, May 26, 2017 3:32 PM
> To: WebAIM Discussion List < = EMAIL ADDRESS REMOVED = >
> Subject: Re: [WebAIM] How useful is the datalist element?
>
> Interesting question - coincidentally, today someone proposed a different
> improvement by allowing grouping:
> https://github.com/w3c/html/issues/932
>
> My sense is that implementation is still immature. And not very good.
> File bugs on browsers...
>
> cheers
>
> On 27/05/17 00:20, Jeremy Echols wrote:
>> It seems like being native HTML, it would be pretty accessible, but in my
>> testing with Firefox and NVDA, it seems like it's... just painful to use.
>> For instance, NVDA doesn't notify me when typeahead options show up, but
>> worse, when I use the down-arrow to start selecting items, it re-reads the
>> input label and aria-describedby text... on every single down-arrow
>> selection.
>>
>> Are there JavaScript typeaheads which are better? Is support for datalist
>> just at such a bad point that it shouldn't be used?
>> >> >> archives at http://webaim.org/discussion/archives
>> >
> --
> Charles McCathie Nevile - standards - Yandex
> = EMAIL ADDRESS REMOVED = - Find more at http://yandex.com
>
> > > http://webaim.org/discussion/archives
> > > > > >


--
Work hard. Have fun. Make history.

From: Jeremy Echols
Date: Tue, May 30 2017 8:34AM
Subject: Re: How useful is the datalist element?
← Previous message | No next message

I'm using it pretty much exactly like the w3schools page (https://www.w3schools.com/tags/tag_datalist.asp), except that I've also got an aria-describedby in there. Their "try it yourself" page seems to have the same problem mine has: too much context is re-read on each use of the down-arrow.

And then there's *zero* Safari support for some reason, which makes it just a standard input box for anybody using Apple products.

I tried out an "accessible combobox" from http://oaa-accessibility.org/examplep/combobox2/, and it actually seems to be worse; NVDA just reads the top two items from the list without notifying me as I scroll through it.

It just seems like the "combo box" concept has been around so long there has to be something that works. But the datalist element falls short. At least for now.

-----Original Message-----
From: WebAIM-Forum [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Birkir R. Gunnarsson
Sent: Sunday, May 28, 2017 9:32 AM
To: WebAIM Discussion List < = EMAIL ADDRESS REMOVED = >
Subject: Re: [WebAIM] How useful is the datalist element?

How are you using the <datalist> element?
Are you using it with the list attribute of the <input> element?


I think it is going to be highly useful but, just like with many HTML5 elements and attributes, the browser and assistive technology support is not quite there yet.
It is a frustrating journey, but if we can keep filing bugs and have this fixed, the payoff in terms of accessibility of complex components is going to be enormous.


On 5/26/17, Jeremy Echols < = EMAIL ADDRESS REMOVED = > wrote:
> This is very disappointing to hear. Thanks for the response.
>
> I don't know enough about the accessibility APIs to know what kind of
> bug to file. I'm assuming the browser is sending the wrong
> information or too much information or something, but is there some
> way to see exactly what NVDA is being told by the browser? And I'm
> not sure what the expected behavior would even be. It seems that the
> concept of a "combo box" hasn't been filled in yet by the w3c "aria practices" authors:
> https://github.com/w3c/aria-practices/issues/31.
>
> -----Original Message-----
> From: WebAIM-Forum [mailto: = EMAIL ADDRESS REMOVED = ] On
> Behalf Of Chaals is Charles McCathie Nevile
> Sent: Friday, May 26, 2017 3:32 PM
> To: WebAIM Discussion List < = EMAIL ADDRESS REMOVED = >
> Subject: Re: [WebAIM] How useful is the datalist element?
>
> Interesting question - coincidentally, today someone proposed a
> different improvement by allowing grouping:
> https://github.com/w3c/html/issues/932
>
> My sense is that implementation is still immature. And not very good.
> File bugs on browsers...
>
> cheers
>
> On 27/05/17 00:20, Jeremy Echols wrote:
>> It seems like being native HTML, it would be pretty accessible, but
>> in my testing with Firefox and NVDA, it seems like it's... just painful to use.
>> For instance, NVDA doesn't notify me when typeahead options show up,
>> but worse, when I use the down-arrow to start selecting items, it
>> re-reads the input label and aria-describedby text... on every single
>> down-arrow selection.
>>
>> Are there JavaScript typeaheads which are better? Is support for
>> datalist just at such a bad point that it shouldn't be used?
>> >> >> archives at http://webaim.org/discussion/archives
>> >
> --
> Charles McCathie Nevile - standards - Yandex
> = EMAIL ADDRESS REMOVED = - Find more at http://yandex.com
>
> > > archives at http://webaim.org/discussion/archives
> > > > archives at http://webaim.org/discussion/archives
> >


--
Work hard. Have fun. Make history.