WebAIM - Web Accessibility In Mind

E-mail List Archives

RE: best screen readers for browser testing

for

From: John Foliot - bytown internet
Date: Jun 19, 2002 8:38AM

> In the case of a government agency which has stated that JAWS (largest
> market share) is the official standard screen reader, then you have no
> choice but to test the 508 requirements with JAWS.
> We also test with IBM Home Page Reader, but not as extensively.

ARGHHH!!!!!!

This sounds dangeously close to: "Since Internet Explorer has an 85% market
share then we have no choice but to test on IE. We also test on Netscape,
but not as extensively." (Never mind that there is also the versioning
issue to contend with...)

Freedom Scientific should be commended for developing and refining a great
piece of Assitive Technology, one which aids the visually impaired use
computers not only for accessing web content, but for all of the other
wonderful things we can do with these machines.

But to say that a web page has been tested in Jaws and it works "fine",
therefore it is <TaDa> accessible, is preposterous. Jaws is a screen
reader, not a SGML parser. Most North American Jaws users will access web
pages using a combination of Jaws and IE, so it would probably stand to
reason that if it "works" in IE, a screen reader would be able to access the
content.

Off list, I was told, as justification to get the latest and most *complete*
software "Without getting into technical details, no matter how compliant
your code is to the HTML 4.01 standard, some of these technologies just will
not interpret and communicate certain tags and information accurately or in
some cases not at all." No contest, but should we then continue to develop
to software instead of standards? Rubbish. Netscape 4.x will not interpret
some of the HTML 4.01 code either (Tabindex, Accesskey, Longdesc, etc.) so
should we not bother testing in Netscape 4.x? Or just not worry about adding
them? (LONGDESC in particular has very dodgy support - should we not bother
then?)

Folks, I'm not advocating NOT testing using AT, but reality check here, just
as you can never replicate every configuration of user agent and operating
platform for the sighted, you cannot replicate every conceivable situation
for the visually impaired (or the mobility impaired, or the cognitively
impaired, or the technology impaired... do you think every blind person out
there can afford to continually upgrade Jaws every time a new version comes
out?).

<mantra>
CODE TO THE STANDARDS, NOT TO THE SOFTWARE
</mantra>
To which I would add the WAI Guidelines as well.

The W3C provides a list of over 35 different "Alternative Web Browsers"
(http://www.w3.org/WAI/References/Browsing), including many text to speech
alternatives. To rely solely on Jaws to replicate the user experience is
then no more logical than relying solely on Internet Explorer on a Windows
system to exclusively replicate the user experience (with a 17" monitor set
at 1024 X 768, or wait, no, a 14" monitor at 800 X 600, or maybe....). But
if you have coded to the standards, then honestly, the software check should
almost be an afterthought (it isn't I know, but it should be...) And so,
Jaws, HPR, Window-eyes, pwWebspeak, Slimware Window Bridge; it shouldn't
matter. True, not all will support everything, but is that really any
different than what the rest of us have to deal with?

As always, JMHO

JF


----
To subscribe, unsubscribe, or view list archives,
visit http://www.webaim.org/discussion/