WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: Al text

for

Number of posts in this thread: 9 (In chronological order)

From: Geof Collis
Date: Tue, Feb 23 2010 7:54AM
Subject: Al text
No previous message | Next message →

Hi All

Just read on another list that using an " " in alt text , for
example alt=" " was acceptable. This goes against anything I've
read so if I'm wrong please let me know.

cheers

Geof



Editor
Accessibility News
www.accessibilitynews.ca
Accessibility News International
www.accessibilitynewsinternational.com

From: Simius Puer
Date: Tue, Feb 23 2010 8:09AM
Subject: Re: Al text
← Previous message | Next message →

Looks like a subtle way to get round failures on automated accessibility
checks.

Can't see it doing any harm, but there is no need for it either...just adds
superfluous code as null alt tags can just be alt=""

From: Geof Collis
Date: Tue, Feb 23 2010 8:18AM
Subject: Re: Al text
← Previous message | Next message →

Thanks Andrew, there argument was to fool checkers, but I dont like
that idea, the checkers I use ask if you've deliberately left it null
so to me it wouldn't be an acceptable practice.

cheers

Geof
At 09:08 AM 2/23/2010, you wrote:
>Looks like a subtle way to get round failures on automated accessibility
>checks.
>
>Can't see it doing any harm, but there is no need for it either...just adds
>superfluous code as null alt tags can just be alt=""
>
>

From: Moore,Michael (DARS)
Date: Tue, Feb 23 2010 9:42AM
Subject: Re: Al text
← Previous message | Next message →

Using alt="&space" can cause some older screen readers to report "graphic space" the standard is to use alt="". If the checker does not recognize alt="" as valid then it is defective. Triggering a manual check when encountering null alt text is not a defect in an automated checker.

Mike Moore

From: Jukka K. Korpela
Date: Tue, Feb 23 2010 10:15AM
Subject: Re: Al text
← Previous message | Next message →

Geof Collis wrote:

> Just read on another list that using an " " in alt text , for
> example alt=" " was acceptable. This goes against anything I've

We must not judge any alt text as unacceptable without knowing the context
and the meaning and role of the image. Even if we cannot imagine acceptable
use, there might be some.

Consider the following fragment of markup:

<img alt="alpha" src="alpha.gif"><img alt="&nbsp;" src="spacer.gif"><img
alt="beta" src="beta.gif">

When images are rendered, this might display as an alpha symbol followed by
spacing of specific width and a beta symbol. A speech browser can be
expected to read it as "alpha beta", as two words. Any speech browser that
cannot treat a no-break space the same way as a space is in need of repair.

The no-break space is effectively a clone of the normal space character,
with additional properties related to line breaking. And this is why here
alt="&nbsp;" might better than alt=" " here, since graphic browser, when not
displaying the images for some reason or another, can be expected to honor
the non-breaking nature of the no-break space. And we can well imagine
reasons why an author would like to keep the words "alpha" and "beta" on the
same line in visual rendering.

I'm not saying that such authoring style is particularly exemplary. But it
isn't absurd or horribly wrong, and assuming that an author has decided to
use small images of symbols to present some expression, like a formula or a
logo, and to use an image for spacing, the question arises what the alt text
should be. What else should it be than the no-break space, given the
premises?

(Of course one can use the no-break space as a character. There is no magic
in using the entity reference &nbsp; instead, except that it is easily
distinguishable from a space in HTML source.)

Even though alt="", alt=" ", and alt="&nbsp;" have all been used to fool
accessibility checkers or to comply with some formal rules or due to mere
misunderstanding, they are (especially the first one, with empty alt text)
have proper uses as well.

--
Yucca, http://www.cs.tut.fi/~jkorpela/

From: ckrugman
Date: Tue, Feb 23 2010 2:03PM
Subject: Re: Al text
← Previous message | Next message →

the problem is that screen readers read all of the garbage codes when that
format is used.
Chuck
----- Original Message -----
From: "Geof Collis" < = EMAIL ADDRESS REMOVED = >
To: "WebAIM Discussion List" < = EMAIL ADDRESS REMOVED = >
Sent: Tuesday, February 23, 2010 6:17 AM
Subject: Re: [WebAIM] Al text


> Thanks Andrew, there argument was to fool checkers, but I dont like
> that idea, the checkers I use ask if you've deliberately left it null
> so to me it wouldn't be an acceptable practice.
>
> cheers
>
> Geof
> At 09:08 AM 2/23/2010, you wrote:
>>Looks like a subtle way to get round failures on automated accessibility
>>checks.
>>
>>Can't see it doing any harm, but there is no need for it either...just
>>adds
>>superfluous code as null alt tags can just be alt=""
>>
>>

From: Jukka K. Korpela
Date: Tue, Feb 23 2010 2:18PM
Subject: Re: Al text
← Previous message | Next message →

= EMAIL ADDRESS REMOVED = wrote:

> the problem is that screen readers read all of the garbage codes when
> that format is used.

Which screen readers? Which "garbage codes"? What "format"? The entity
reference &nbsp; is just a notation that denotes the no-break space
character " ". If your screen reader doesn't grok it, then it needs to be
fixed. The no-break space character has been used on web pages since the
early days, and if your software chokes on it, it might well choke on "!" as
well. Should we therefore stop using exclamation marks?

--
Yucca, http://www.cs.tut.fi/~jkorpela/

From: Geof Collis
Date: Tue, Feb 23 2010 2:24PM
Subject: Re: Al text
← Previous message | Next message →

I did an experiment using alte="", alt=" " and alt="&nbsp;" on 2
computers, one with JAWS 10 and IE8.0 and Firefox 3., and the other
JAWS 6.0 and IE 6.0

On JAWS 6.0 none of them was visible but with JAWS 10 on IE 8.0 and
Firefox the alt="&nbsp;" was visible, so I dont think it is a good idea.

cheers

Geof

At 03:18 PM 2/23/2010, you wrote:
> = EMAIL ADDRESS REMOVED = wrote:
>
> > the problem is that screen readers read all of the garbage codes when
> > that format is used.
>
>Which screen readers? Which "garbage codes"? What "format"? The entity
>reference &nbsp; is just a notation that denotes the no-break space
>character " ". If your screen reader doesn't grok it, then it needs to be
>fixed. The no-break space character has been used on web pages since the
>early days, and if your software chokes on it, it might well choke on "!" as
>well. Should we therefore stop using exclamation marks?
>
>--
>Yucca, http://www.cs.tut.fi/~jkorpela/
>
>

From: ckrugman
Date: Tue, Feb 23 2010 3:42PM
Subject: Re: Al text
← Previous message | No next message

In JAWS it will repeated read the "&NBST" when it is present on many web
sites. This is distracting and wastes time.
Chuck
----- Original Message -----
From: "Jukka K. Korpela" < = EMAIL ADDRESS REMOVED = >
To: "WebAIM Discussion List" < = EMAIL ADDRESS REMOVED = >
Sent: Tuesday, February 23, 2010 12:18 PM
Subject: Re: [WebAIM] Al text


> = EMAIL ADDRESS REMOVED = wrote:
>
>> the problem is that screen readers read all of the garbage codes when
>> that format is used.
>
> Which screen readers? Which "garbage codes"? What "format"? The entity
> reference &nbsp; is just a notation that denotes the no-break space
> character " ". If your screen reader doesn't grok it, then it needs to be
> fixed. The no-break space character has been used on web pages since the
> early days, and if your software chokes on it, it might well choke on "!"
> as
> well. Should we therefore stop using exclamation marks?
>
> --
> Yucca, http://www.cs.tut.fi/~jkorpela/
>
>