W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2002

RE: JAWS and name attribute for IMG

From: john_slatin <john_slatin@forum.utexas.edu>
Date: Thu, 2 May 2002 10:34:46 -0500
Message-ID: <6AC4E20EED49D411941400D0B77E52F006C06D3C@forum.cc.utexas.edu>
To: "'Chris O'Kennon'" <chris@vipnet.org>, "'Jim Ley'" <jim@jibbering.com>, john_slatin <john_slatin@forum.utexas.edu>, "'WCAG (E-mail)'" <w3c-wai-gl@w3.org>
Jim, thanks for pointing out that I forgot to mention which browser I was

I was using IE 5.5 with JAWS 4.01 under Win98.  The same behavior occurred
with IE 6.0 and JAWS 4.02 under WinXP.

I can think of several reasons why it's *not* appropriate for JAWS to report
the name attribute on IMG elements.

Valid HTML specifically mandates that the ALT attribute be present for *all*
IMG elements.  Good practice also requires that ALT be set to null (ALT="")
for such things as spacer Gis and other images that do not carry content,
especially when those images are repeated and/or when encountering them
would distract or confuse someone using a screen reader, talking browser,
refreshable Braille, or text-only display.

If developers then use the name attribute for internal management (e.g., to
keep track of which images go with which Javascript calls, as was the case
on the page where I first encountered this), and if AT then *speaks* (or
otherwise renders) the content of the name attribute which users were not
intended to hear or encounter, the required technique of setting ALT="" is
defeated, and *no one* benefits from the defeat!

Chris, in response to your question about the way the ALT attribute is used
on the Virginia site, I'd say "no"-- an ALT text that's merely an
identification number doesn't convey any meaning to the user, and the
purpose of the ALT attribute is to provide a text *equivalent* for the
functionality of the image.  If the function of the image is to create blank
space for layout (for example), then the equivalent functionality for speech
output is sielcne.

How or whether users who are blind/have low vision hear about images that
have no ALT text depends on how they configure their screen readers, at
least for JAWS: if I tell JAWS to report "Only tagged graphics," it will
ignore images without ALT attributes in exactly the same way that it will
ignore images with ALT="".  That is, in both cases it will act as though
there is no image present.  "Only tagged images" is the default setting;
therefore, if there *is* ALT text there, JAWS will report it whether it's
meaningful or not.  It should either be meaningful or be empty00 that is,

Sorry that was so long!

-----Original Message-----
From: Chris O'Kennon [mailto:chris@vipnet.org] 
Sent: Thursday, May 02, 2002 8:16 am
To: 'Jim Ley'; john_slatin; WCAG (E-mail)
Subject: RE: JAWS and name attribute for IMG

This brings up a discussion I recently had at a conference in New Orleans.
We had a visually impaired presenter tell us that she prefers that images
used solely for layout and convey no information be identified with ALT="".
I maintain the Commonwealth of Virginia portal, and we use ALT="#" (with the
first image identifying what the # stands for).  The reason we do this is
because there didn't seem to be any way for a blind user to tell the
difference between an image with an empty ALT tag and an image someone
FORGOT to ALT tag.  Has there been any discussion (and I apologize if there
has been, as I'm a newbie here) on how to best address this?  Does the NAME
attribute work on all browsers?

Chris O'Kennon
Commonwealth of Virginia Webmaster/
VIPNet Portal Architect

"This had better work." 
	-Grand Moff Tarkin

> -----Original Message-----
> From: Jim Ley [mailto:jim@jibbering.com]
> Sent: Thursday, May 02, 2002 8:19 AM
> To: john_slatin; WCAG (E-mail)
> Subject: Re: JAWS and name attribute for IMG
> "john_slatin" <john_slatin@forum.utexas.edu>
> > I just discovered something surprising (at least to me):
> >
> > For IMG elements, JAWS 4.01 reports the *name* attribute when  
> > ALT="" and there is no TITLE attribute.
> I think this would make a lot of sense in the case where
> there was no ALT
> attribute at all. You don't say which browser you're using 
> with JAWS but
> it's possible that the browser doesn't expose the difference between
> ALT="" and no ALT in the DOM, in which case it would still 
> need to do the
> repair activity.
> Jim.
Received on Thursday, 2 May 2002 11:34:50 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:41 UTC