WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: Re: Foreign:Re: Screenreader support for title....

for

Number of posts in this thread: 4 (In chronological order)

From: Moore, Michael
Date: Tue, Jun 03 2008 7:10AM
Subject: Re: Foreign:Re: Screenreader support for title....
No previous message | Next message →

Paul,

In terms of JAWS, the TITLE attribute is supported, however the end-user
must turn that option on (at least in JAWS V7.X), it is not
automatically read out-of-the-box.

Jennison

Close Jennison, but not entirely accurate. If the label attribute is
absent but the title attribute is present JAWS will read the title by
default. This has been the case since at least JAWS 4. Window Eyes
demonstrates similar behavior. In JAWS, the verbosity settings can be
set to favor the title over the label, to read both, or to read which
ever is longer. This selection appears to be working correctly in JAWS
9 but did not work in JAWS 6. I am not sure exactly when the problem
was resolved. In JAWS 6, the screen reader stubbornly refused to read
the title if a properly marked up label was present. From a practical
standpoint this was not an issue, assuming that the labels were marked
up correctly and contained appropriate text.

Mike

From: Paul Collins
Date: Tue, Jun 03 2008 7:20AM
Subject: Re: Screenreader support for title....
← Previous message | Next message →

Thanks for your replies everyone and thanks for the links to the
research by Steve.

I probably should have made that more clear; I meant using a title on
a HREF, although it is interesting to see the amount of support for it
on form elements.

I've been ignoring using it for a while due to what I assumed was a
lack of support; and putting all valid information in the ALT tag when
it is a HREF around and image.

It would be good if assistive software came on board and had more
support for the title attribute. Another way around it would be HTML
support for the ALT attribute on the HREF, so you could make your
non-descriptive text links more descriptive.

Or, maybe that's going to confuse things more :)

Thanks again,
Paul




2008/6/3 Jukka K. Korpela < = EMAIL ADDRESS REMOVED = >:
> Paul Collins wrote:
>
>> Just wondering what the latest is on this? The title never used to be
>> read out by JAWS, IBM and the like, although that was supposed to be
>> changing. Does anyone have the latest on which popular screenreaders
>> read out both the title and alt attribute?
>
> For some odd reason, the documents I found with simple searches look
> rather old and sketchy. Moreover, the situation surely varies by
> element. For example,
> http://www.paciellogroup.com/resources/articles/WE05/forms.html
> is from year 2005 and describes even JAWS 4.02 as supporting the title
> element for several elements (though the <img> element, which you might
> be primarily interested in, is "not tested").
>
> But _some_ optional support (as a user-selectable option) has existed
> for several years.
>
> However, I think the issue is mostly pointless. Why would it matter,
> when we _know_ that the great majority of users won't access the
> information in title attributes anyway? A "normal" user might
> accidentally hit some "titled" element with the mouse and then notice
> the tiny text in a small popup, but mostly nobody sees what you put in a
> title attribute.
>
> If some information needs explanation, explain it in normal content.
> Putting the information in an attribute may look modern and advanced,
> but it's like publishing a book with annotations printed in invisible
> ink that becomes visible in fluorescent light.
>
> Jukka K. Korpela ("Yucca")
> http://www.cs.tut.fi/~jkorpela/
>
>

From: Aaron Cannon
Date: Tue, Jun 03 2008 8:30AM
Subject: Re: Screenreader support for title....
← Previous message | Next message →

I tried to figure out a definitive answer a few days ago using Jaws 8.0. I have not changed any of the defaults that effect this. Here is what I found:
The title attribute is read for links when there is no link text. E.G. <a title="bla" href="..."></a>
The title attribute is read for images when there is no alt attribute. (Note that it is not read if the alt attribute is present but empty.)
The title attribute on form fields is only read when there is no label.

Aaron


>>> "ben morrison" < = EMAIL ADDRESS REMOVED = > 6/3/2008 6:23 AM >>>

On Tue, Jun 3, 2008 at 12:18 PM, Jennison Mark Asuncion
< = EMAIL ADDRESS REMOVED = > wrote:
> Paul,
>
> In terms of JAWS, the TITLE attribute is supported, however the
> end-user must turn that option on (at least in JAWS V7.X), it is not
> automatically read out-of-the-box.

I'm not sure about latest versions, but I think it would work in forms
mode/form elements but not on any other elements say <a> unless
'verbosity' settings were changed.

I just don't rely on the title attribute.

ben
--
Ben Morrison

From: Steve Green
Date: Tue, Jun 03 2008 11:20AM
Subject: Re: Screenreader support for title....
← Previous message | No next message

-----Original Message-----
From: = EMAIL ADDRESS REMOVED =
[mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Paul Collins
Sent: 03 June 2008 14:13
To: WebAIM Discussion List
Subject: Re: [WebAIM] Screenreader support for title....

Thanks for your replies everyone and thanks for the links to the research by
Steve.

I probably should have made that more clear; I meant using a title on a
HREF, although it is interesting to see the amount of support for it on form
elements.

I've been ignoring using it for a while due to what I assumed was a lack of
support; and putting all valid information in the ALT tag when it is a HREF
around and image.

It would be good if assistive software came on board and had more support
for the title attribute. Another way around it would be HTML support for the
ALT attribute on the HREF, so you could make your non-descriptive text links
more descriptive.

Or, maybe that's going to confuse things more :)

Thanks again,
Paul



Assistive technologies are not the problem. The problem is that you are
trying to present information by means of a non-standard method of user
interaction. The basic interaction model is that stuff is displayed on
screen and you click something to view more stuff. Hovering over an element
to obtain more information is not an expected means of interaction.

The standards do not define how support for 'title' attributes should be
implemented in user agents. The most common ones implement it by hovering
over an element, but few users are aware of this and some cannot interact
this way such as people who use keyboard navigation or voice recognition
software.

Steve