WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: WCAG

for

Number of posts in this thread: 21 (In chronological order)

From: Geof Collis
Date: Tue, Mar 02 2010 11:18AM
Subject: WCAG
No previous message | Next message →

Hi All

Is there a guideline that refers to the number of characters in alt
text before it requires another method? I cant seem to find one.

cheers

Geof

Administrator
Coalition of Ontario Accessibility Advisory Committees (COAAC) Website
www.coaac.ca

From: Simius Puer
Date: Tue, Mar 02 2010 11:33AM
Subject: Re: WCAG
← Previous message | Next message →

I have it in my head that there is no actual specification that limits this,
although the figure of 125 characters keep popping into my head relating to
some guideline...not on my PC otherwise I could check my archives for this
info.

From: Patrick H. Lauke
Date: Tue, Mar 02 2010 11:36AM
Subject: Re: WCAG
← Previous message | Next message →

On 02/03/2010 17:17, Geof Collis wrote:
> Hi All
>
> Is there a guideline that refers to the number of characters in alt
> text before it requires another method? I cant seem to find one.

As that's tech specific, it wouldn't be in the guidelines. Look in the
techniques http://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/G94

P
--
Patrick H. Lauke

From: Patrick H. Lauke
Date: Tue, Mar 02 2010 11:39AM
Subject: Re: WCAG
← Previous message | Next message →

On 02/03/2010 17:34, Simius Puer wrote:
> I have it in my head that there is no actual specification that limits this,
> although the figure of 125 characters keep popping into my head relating to
> some guideline...not on my PC otherwise I could check my archives for this
> info.

That seems to be related to how JAWS specifically handles ALT.

http://www.washington.edu/accessit/articles?1257

--
Patrick H. Lauke

From: Jared Smith
Date: Tue, Mar 02 2010 11:42AM
Subject: Re: WCAG
← Previous message | Next message →

On Tue, Mar 2, 2010 at 10:35 AM, Patrick H. Lauke
< = EMAIL ADDRESS REMOVED = > wrote:

> As that's tech specific, it wouldn't be in the guidelines. Look in the
> techniques http://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/G94

WCAG 2.0 uses "short text alternative" and "long description"
extensively in the techniques documents, but as far as I know, they
never define either term. I'm unfamiliar with an exact character value
defined or recommended in WCAG 2.0, but I distinctly remember that at
one point the suggestion was "around 100 characters". This is, of
course, a rough guideline - certainly some images should have alt
values that are longer than 100 characters, but I think this is a good
point at which you should seriously consider either making the alt
text more shorter or providing a longer description.

Jared Smith
WebAIM

From: Karl Groves
Date: Tue, Mar 02 2010 11:51AM
Subject: Re: WCAG
← Previous message | Next message →

>
> > As that's tech specific, it wouldn't be in the guidelines. Look in
> the
> > techniques http://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/G94
>
> WCAG 2.0 uses "short text alternative" and "long description"
> extensively in the techniques documents, but as far as I know, they
> never define either term. I'm unfamiliar with an exact character value
> defined or recommended in WCAG 2.0, but I distinctly remember that at
> one point the suggestion was "around 100 characters". This is, of
> course, a rough guideline - certainly some images should have alt
> values that are longer than 100 characters, but I think this is a good
> point at which you should seriously consider either making the alt
> text more shorter or providing a longer description.
>


I think it bears mentioning that the "Techniques" document is informative,
rather than normative. They supply guidance on how to conform to the
guidelines. The techniques may or may not be applicable to your specific
situation and some may eventually reach a state of obsolescence by future
developments in technology and/ or AT.

As Jared pointed out "around 100 characters" is a good rule of thumb.
Whatever approach you decide should you come to your own conclusions should
be predicated of course on AT and user agent support.

Karl

From: Patrick H. Lauke
Date: Tue, Mar 02 2010 11:54AM
Subject: Re: WCAG
← Previous message | Next message →

On 02/03/2010 17:42, Jared Smith wrote:
> WCAG 2.0 uses "short text alternative" and "long description"
> extensively in the techniques documents, but as far as I know, they
> never define either term. I'm unfamiliar with an exact character value
> defined or recommended in WCAG 2.0, but I distinctly remember that at
> one point the suggestion was "around 100 characters". This is, of
> course, a rough guideline - certainly some images should have alt
> values that are longer than 100 characters, but I think this is a good
> point at which you should seriously consider either making the alt
> text more shorter or providing a longer description.

I took the vague "short" and "long" terminology to be more about common
sense. They don't explicitly define what is long or short, as it's
heavily dependent on language (compare the amount of info that, say, 100
chars in japanese can convey vs german).

(Although in 1.4.8, they do explictly mention char counts with regards
to line length, which i still don't quite agree with as it seems
arbitrary: "Width is no more than 80 characters or glyphs (40 if CJK).")

I got the impression from the thread starter that it was more about an
absolute maximum in terms of "in HTML, using ALT attribute, before it
causes problems in specific browsers / with specific AT", which is
dependent on the tech.

P
--
Patrick H. Lauke

From: Simius Puer
Date: Tue, Mar 02 2010 11:57AM
Subject: Re: WCAG
← Previous message | Next message →

Yes, I was quite disappointed to see WCAG 2.0 failing to include limits or
even guidance on the length, either in the guidelines or the tech.

And Patrick, the 125 was a limit in older versions of Jaws but I don't think
it is now. The figure of 125 chars didn't come from there...I remember
reading something about this a few years back. It was not W3C/WCAG but it
was most certainly accessibility orientated...and as Jared says, there
really needs to be some common sense regarding the method used.

From: Patrick Burke
Date: Tue, Mar 02 2010 12:12PM
Subject: Re: WCAG
← Previous message | Next message →

At 09:51 AM 3/2/2010, Simius Puer wrote:
>Yes, I was quite disappointed to see WCAG 2.0 failing to include limits or
>even guidance on the length, either in the guidelines or the tech.
>
>And Patrick, the 125 was a limit in older versions of Jaws but I don't think
>it is now. The figure of 125 chars didn't come from there...I remember
>reading something about this a few years back. It was not W3C/WCAG but it
>was most certainly accessibility orientated...and as Jared says, there
>really needs to be some common sense regarding the method used.

These days I think we can all agree on 140 chars as the standard unit
of measurement to determine when the Too-Much-Data Point has been reached.

So ALT text should be no more than 1 tweet in length.

LOL, as they say ...


A Different Patrick
--
Patrick J. Burke

Coordinator
UCLA Disabilities &
Computing Program

Phone: 310 206-6004
E-mail: burke <at> ucla. edu

From: Geof Collis
Date: Tue, Mar 02 2010 12:51PM
Subject: Re: WCAG
← Previous message | Next message →

One of the Validators I use says it is 150 characters and gives an
error if it is over.

cheers

Geof

At 01:11 PM 3/2/2010, you wrote:
>At 09:51 AM 3/2/2010, Simius Puer wrote:
> >Yes, I was quite disappointed to see WCAG 2.0 failing to include limits or
> >even guidance on the length, either in the guidelines or the tech.
> >
> >And Patrick, the 125 was a limit in older versions of Jaws but I don't think
> >it is now. The figure of 125 chars didn't come from there...I remember
> >reading something about this a few years back. It was not W3C/WCAG but it
> >was most certainly accessibility orientated...and as Jared says, there
> >really needs to be some common sense regarding the method used.
>
>These days I think we can all agree on 140 chars as the standard unit
>of measurement to determine when the Too-Much-Data Point has been reached.
>
>So ALT text should be no more than 1 tweet in length.
>
>LOL, as they say ...
>
>
>A Different Patrick
>--
>Patrick J. Burke
>
>Coordinator
>UCLA Disabilities &
>Computing Program
>
>Phone: 310 206-6004
>E-mail: burke <at> ucla. edu
>
>

From: Jared Smith
Date: Tue, Mar 02 2010 1:00PM
Subject: Re: WCAG
← Previous message | Next message →

On Tue, Mar 2, 2010 at 11:50 AM, Geof Collis < = EMAIL ADDRESS REMOVED = > wrote:
> One of the Validators I use says it is 150 characters and gives an
> error if it is over.

Then the validator is simply applying an arbitrary definition of
accessibility. Any validator result that flags something like this as
an error should be viewed with great caution. In this case, if the
alternative text is appropriate and still over 150 characters, then
simply ignore the error and carry on. However, the instances of
appropriate alternative text being over 150 characters are quite rare.
Luckily we, as intelligent humans, rather than tools get to decide
what is and is not accessible.

Jared

From: Geof Collis
Date: Tue, Mar 02 2010 1:33PM
Subject: Re: WCAG
← Previous message | Next message →

I think I'll go with it as it sounds correct to me. I've always
understood that 150 characters was the max


cheers

Geof

At 02:01 PM 3/2/2010, you wrote:
>On Tue, Mar 2, 2010 at 11:50 AM, Geof Collis < = EMAIL ADDRESS REMOVED = > wrote:
> > One of the Validators I use says it is 150 characters and gives an
> > error if it is over.
>
>Then the validator is simply applying an arbitrary definition of
>accessibility. Any validator result that flags something like this as
>an error should be viewed with great caution. In this case, if the
>alternative text is appropriate and still over 150 characters, then
>simply ignore the error and carry on. However, the instances of
>appropriate alternative text being over 150 characters are quite rare.
>Luckily we, as intelligent humans, rather than tools get to decide
>what is and is not accessible.
>
>Jared
>

From: Moore,Michael (DARS)
Date: Tue, Mar 02 2010 1:42PM
Subject: Re: WCAG
← Previous message | Next message →

Newer releases of JFW support reading longer alt texts which can be convenient for complex diagrams and graphs.

Mike Moore


From: Jared Smith
Date: Tue, Mar 02 2010 1:54PM
Subject: Re: WCAG
← Previous message | Next message →

On Tue, Mar 2, 2010 at 12:34 PM, Geof Collis < = EMAIL ADDRESS REMOVED = > wrote:
> I think I'll go with it as it sounds correct to me. I've always
> understood that 150 characters was the max

You're missing the point entirely. There is no max!!! Setting any
value to alternative text length will be entirely arbitrary. It would
be like saying any book of fiction over 150 pages is automatically a
novel and everything less is a short story. Absurd!

Alternative text should convey the content and function of an image,
and do so accurately and succinctly. Some images necessitate very long
alternative text. Most require a word or two. I think 100 characters
is a good threshold at which one should determine whether the
alternative for an image can adequately be conveyed through the alt
attribute. In some cases, this will be perfectly appropriate - maybe
even at 1000 or even 10000 characters (especially for images that
present lots of text). But in most cases, the fact that the alt value
is naturally becoming lengthy is an indication that the alternative
might be better presented in some other way, such as in adjacent text
or another page via link and (optionally) the longdesc attribute.

Jared

From: Christine Peterson
Date: Tue, Mar 02 2010 1:57PM
Subject: Re: WCAG
← Previous message | Next message →

A side note. I was just looking through the draft of HTML 5 and noticed that one of the attributes they are taking out is the longdesc. I'm not familiar enough with this process to know if there will be a replacement or if there will not be a way to include long descriptions of images.

Granted, the standards process isn't done yet, so this may not be the final say . . .

http://www.w3.org/TR/html5-diff/#absent-attributes


Christine Peterson
= EMAIL ADDRESS REMOVED =

Assess>>Act>>Advance, the Amigos 2010 Annual Member Conference - online
May 5 and 6, 2010    http://conference.amigos.org/

From: Geof Collis
Date: Tue, Mar 02 2010 2:00PM
Subject: Re: WCAG
← Previous message | Next message →

I reserve the right to be absurd then.

cheers

Geof

At 02:54 PM 3/2/2010, you wrote:
>On Tue, Mar 2, 2010 at 12:34 PM, Geof Collis < = EMAIL ADDRESS REMOVED = > wrote:
> > I think I'll go with it as it sounds correct to me. I've always
> > understood that 150 characters was the max
>
>You're missing the point entirely. There is no max!!! Setting any
>value to alternative text length will be entirely arbitrary. It would
>be like saying any book of fiction over 150 pages is automatically a
>novel and everything less is a short story. Absurd!
>
>Alternative text should convey the content and function of an image,
>and do so accurately and succinctly. Some images necessitate very long
>alternative text. Most require a word or two. I think 100 characters
>is a good threshold at which one should determine whether the
>alternative for an image can adequately be conveyed through the alt
>attribute. In some cases, this will be perfectly appropriate - maybe
>even at 1000 or even 10000 characters (especially for images that
>present lots of text). But in most cases, the fact that the alt value
>is naturally becoming lengthy is an indication that the alternative
>might be better presented in some other way, such as in adjacent text
>or another page via link and (optionally) the longdesc attribute.
>
>Jared
>

From: Moore,Michael (DARS)
Date: Tue, Mar 02 2010 2:03PM
Subject: Re: WCAG
← Previous message | Next message →

I agree that the example is too long. "Download assistive technology software for individuals with physical disabilities" is what I would use, but the marketing people would probably insist that I add the brand name. Since it is a link, the destination/function of the link is the appropriate alt text.

Mike Moore

From: Geof Collis
Date: Tue, Mar 02 2010 2:09PM
Subject: Re: WCAG
← Previous message | Next message →

That would be understandable but this question has to do with the alt
text I mentioned the other day, it was just a logo/link with the
following alt text:
This Web Accessibility icon serves as a link to download eSSENTIAL
Accessibility assistive technology software for individuals with
physical disabilities.


cheers

Geof




At 02:41 PM 3/2/2010, you wrote:
>Newer releases of JFW support reading longer alt texts which can be
>convenient for complex diagrams and graphs.
>
>Mike Moore
>
>
>

From: Geof Collis
Date: Tue, Mar 02 2010 2:12PM
Subject: Re: WCAG
← Previous message | Next message →

After consulting with you good peeps I adjusted it last week and the
error went away. :O)

cheers

Geof

At 02:59 PM 3/2/2010, you wrote:
>I agree that the example is too long. "Download assistive technology
>software for individuals with physical disabilities" is what I would
>use, but the marketing people would probably insist that I add the
>brand name. Since it is a link, the destination/function of the link
>is the appropriate alt text.
>
>Mike Moore
>
>

From: Hoffman, Allen
Date: Tue, Mar 02 2010 5:15PM
Subject: Re: WCAG
← Previous message | Next message →

It would be in the HTML specification.
Would be a good test for an HTML validate.
Its not a WCAG specification.

Make the file with alt tags of various lengths and find out what you get
in return from the validator.
Put a whole page inside an alt tag.



From: Jennison Mark Asuncion
Date: Tue, Mar 02 2010 9:15PM
Subject: Re: WCAG
← Previous message | No next message

Hi there,

Perhaps it is because I am not necessarily an audio learner, but as a
screen reader user, shorter is better re the ALT text. If it's too long,
especially if it is describing something complex, I'll more than
likely lose the intended meaning. But that's just me. I was the same way
back in school. Describing a complex Calculus diagram to me in words was
lost on me, it had to be drawn out using raised lines for me to really
"get it." I do know that there are other peers of mine who are blind who
are "wired" in such a way that you can describe something to them that may
be complex, and they get it. Envious.

That would be an interesting study for someone to take on re how much
someone who can't see really gets out of different styles of ALT text .

Jennison

Jennison Asuncion
Co-Director, Adaptech Research Network http://www.adaptech.org
LinkedIn at http://www.linkedin.com/in/jennison