W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2004

Re: Screen readers - usage stats?

From: David Poehlman <poehlman1@comcast.net>
Date: Wed, 14 Apr 2004 16:13:19 -0400
Message-ID: <001901c4225c$eecceec0$6601a8c0@hands>
To: "Ian Anderson" <lists@zstudio.co.uk>, <w3c-wai-ig@w3.org>

Yes, Most often, the issues I have seen are poorer rendering of something by
jaws than window eyes.  this is with jaws 4.51 and window eyes 4.5 but I am
sure there are other issues with other versions.  I agree that market
figures would be interesting, and can tell you that from what I have seen,
jaws outstrips all the others by quite a wide margin and has for some time
but I have also seen issues of rendering dealt with in ways that they are
resolved at least for a particular version of one screen reader over another
as far as jaws and window eyes are concerned.  Since both jaws and window
eyes rely on ie to pre-render their pre-rendering, I'd suspect that the
rendering differences you find have more to do with what may have been
determined by the screen reader developpers as what is prefered by their
users or that are restricted by their particular operating environment visa
vie msaa vs non msaa parsing.  and as you say, the defaults for the screen
readers are different and can cause further percieved issues.  For instance,
if you tell jaws to read all graphics, if alt="" or alt=" " are present as
alt attributes, they will not have the desired effect of null reading.  I
think others may handle this differently.

What needs to be kept in mind is that the users of a particular screen
reader usually expect certain behaviors and know how to deal with them and
have set up the environment according to their needs so if you code
something that you think has solved an issue for a particular screen reader
and that screen reader is set to render differently by a user you may still
get a complaint.  This is what makes coding for screen readers difficult in
addition to the fact that you may assume that a jaws user is using ie bu in
fact, that user might be using netscape 4.075.

My personal preferences in jaws departs a good bit from the defaults but
when I encounter sites that are well coded, I have no accessibility issues
with them.

----- Original Message ----- 
From: "Ian Anderson" <lists@zstudio.co.uk>
To: <w3c-wai-ig@w3.org>
Sent: Wednesday, April 14, 2004 3:49 PM
Subject: Re: Screen readers - usage stats?

> What compelling reason could there be to have stats of this type?  I can
> think of some that have nothing to do with the web but none that do.

I'd be interested to see information on screen reader market share. I'm
consulting with a UK Internet bank presently, and we are testing in
Window-Eyes 4.5, JAWS 5.0, JAWS 4.5 and IBM HomePage Reader 3.0. You get
problems with the default settings in each configuration, and slightly
different HTML code is required to make site features work well in each one.
Often there are conflicts, where the code that works well in JAWS causes a
rendering issue in Window-Eyes, or vice versa.

While I don't believe in trying to design the accessible experience (because
it's impossible :), where there are conflicts  I tend to lean towards the
JAWS 4.5 rendering if there are choices to be made, because I estimate that
this is the largest user base. In six months time, that may not be the case,
and this is why information about market share would be useful.

It's a situation analogous to the old dilemmas facing web designers when we
cared about NN4 versus IE4 rendering of pages. That was solved as the modern
browsers began supporting the standards more consistently, so that the
latest versions of IE, NN, Opera, Mozilla, and so on all render pages 95%
consistently or better.

Since the screen readers differ so much in their rendering of standard HTML
constructs, there are necessary decisions to be made. If there were no
conflicts, I wouldn't care so much about market share.

Perhaps in a few years the manufacturers of screen readers will get their
act together, but in the meantime the reality is that someone will always
get a poorer experience, and web designers are back to juggling the numbers
again to decide which way to jump for each design choice.

Incidentally, I'm talking about screen readers here because they have
rendering differences that affect the user experience in significant ways,
and are highly affected by small differences in the coding of a page. Screen
magnifiers typically are not, for example, so that's why I'm talking about
screen readers in particular.

Anyone else wrestled with screen reader rendering differences, and have any


Ian Anderson
Received on Wednesday, 14 April 2004 16:13:34 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:28 UTC