W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2002

Re: Creative (?) solution for redundant links

From: Patrick Burke <burke@ucla.edu>
Date: Tue, 12 Mar 2002 13:13:37 -0800
Message-Id: <>
To: w3c-wai-ig@w3.org
I haven't had this conversation with the Jaws folks (at least not recently, 
so disregard me if you wish). However, colleagues who have talked to them 
say that it was indeed a deliberate choice. The idea being that Jaws 
developers didn't trust sites to use ALT correctly (which may have been a 
valid concern a couple of years ago). So empty ALTs or single characters or 
strings of repeated characters get thrown out & just the destination 
graphic filename is shown.

Now that the use of ALT="" etc. is (somewhat?) more widely understood, it 
would be much better for Jaws to stick with displaying ALT (or TITLE) as is.

So, I agree with David Poehlman:
At 11:00 AM 3/12/2002, David Poehlman wrote:
>this issue is being worked.  If I ask for all images, it should tell me
>that there is nothing in the alt tag not give me the file name.
If a solution is in process, then hurray! ... Although I would add that the 
original Jaws behavior should still be an option, for those cases where ALT 
is in fact misused. (Then the user won't have to slog through source code 
to find the image name. ... Actually I have encountered an interesting 
example of this on a site with a library of emoticons for a message board. 
When the emoticons had no ALT attributes, I could still read the filenames, 
which are descriptive enough. Now that they have added alt="icon" to all of 
them, I have no way to tell them apart. Of course the real solution is to 
have good ALT text, but a full arsenal of screen reader tricks is always 
good to have.)


At 01:57 PM 3/11/2002, David Woolley wrote:
> >
> > Jaws has a bug that forces the file name of the image to be rendered
>Surely that is a feature to work round sites which provided null alt
>text when they should have provided real alt text.
Received on Tuesday, 12 March 2002 16:09:02 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:36:07 UTC