Re: Alt Text (Was: Re: Bobby inaccuracy?)

At 01:08 PM 2002-01-15 , Liz Roberts wrote:
>> Hope this clarifies things a bit. While the alt attribute is widely
>> misunderstood, its intended use is not controversial.
>
>Every time a question about alt attributes appears on the list, a plethora
>of answers--many contradictory--follow.
>
>These are poorly written, but here's what I've gleaned from the list to do:
>
>1. the alt attribute should be a text _equivalent_; convey the same
>information/serve the same purpose as the image.
>
>2. place a period after appropriate alt text to stop the flow of text
>(obviously the exception is when this stoppage is undesirable).
>
>3. use alt="" for spacer images or similar types of graphics/etc that don't
>have specific meaning.  (I didn't say this very well, sorry.)

.. or even non-specific meaning.

You said that fine.  When in doubt, use hypertext.  Find 

 Side-By-Side
 <http://jimthatcher.com/sidebyside.htm>http://jimthatcher.com/sidebyside.htm

.. and refer to it on this point.  See government reference below as well.

>
>But I've still got questions, and I'm sure my three previous (personal)
>"rules" will be disputed (please do so).
>
>- Is alt="" or alt=" " better?  (There was a debate on this a while back; I
>got lost.)

Have you honed your Google skills to find these things?

>- If I describe a double helix in a document and accompany this document
>with an image of a double helix, what alt text should the image have?  What
>about if it's a document about beach vacation properties and it has images
>of the water and people playing in the sand?

Seriously, one cannot answer this question in the absence of the actual
discussion in the document.  It depends.  It can be good to have a one-liner
for the double helix simply as a summary of the _looong_ description in the
main text.  Is there a figure caption in the page layout?  etc.  But an action
caption such as "two very long chains of nucleotides paired and twisted
together in a spiral" could make sense for the DNA structure if that level of
summary is not in the layout.

>
>- In that same vein, do these or similar "window dressing" images merit a
>longdesc?  (And how well is longdesc supported by assitive technologies?)
>

Longdesc: not well.

Description: get _someone else_ to look at the page and ask them so they have
to respond orally, "what is this [the image] doing here?"  That will get you
close to what to cover and how deeply.

See also

affective messaging and effective mode-crossing (desc example)
<http://lists.w3.org/Archives/Public/wai-tech-comments/2001Aug/0001.html>ht
tp://lists.w3.org/Archives/Public/wai-tech-comments/2001Aug/0001.html

>Additionally, I've had ADA compliance officers who work for some of the
>organizations we do work for tell me not to use alt=" " and _always_ specify
>a meaning.  So if your header graphic has three pieces, have "My Great
>Homepage (graphic 1 of 3).", "My Great Homepage (graphic 2 of 3).", etc.  I
>can't quite deal with that one as I know I wouldn't want to hear this junk.
>Any idea why an ADA guru would recommend that? (When I challenged the idea,
>the answer was "because of how assitive technologies work: if you have alt
>equals blank, it will read the image name.")

Compliance officers have to push people all day.  So sometimes they hedge by
overcompensating.

Do refer them to the Government statements on Section 508.  These two laws are
not synonymous, but any court that wouldn't consider the 508 precedents in this
area would be out to lunch.

<http://www.access-board.gov/sec508/guide/1194.22.htm#(a>http://www.access-
board.gov/sec508/guide/1194.22.htm#(a)

Reading the file name is not "how assistive technologies work."  It is how some
assistive technologies work.  It is good, but ignoring the item is better. 
Compare with UAAG checkpoints 
2.7 and 2.8, where always putting something in there is good but supporting
auto-hiding of instances of ALT="" is better.  This was arrived at after much
debate across different segments of stakeholders.

>
>FWIW, I would second Charles' opinion to just use "Sheffield City Council."
>The purpose this graphic serves here is as your identifying title.
>
>Finally, a quick apology: I feel this might not be the best place for these
>questions, but this is--ironically--the only accessibility list I've found
>with any traffic!

Just my personal 'participation' standards, but the way you framed this post
counts in my book as 'value added' and 'a good model to follow.'  "This is what
I think I've learned, and this is what I am not so sure about" is a
far-above-median quality level of professionalism for unmoderated list email,
IMHO.

Al
>
>Thanks,
>Liz
>

Received on Tuesday, 15 January 2002 14:26:48 UTC