Re: Screen readers - usage stats?

Ian and all,
The users of a t need to fight their own battles if you follow best
practices as you point out near the end of your note below.  The A T will
change much faster than formats change, so the best you can do is to put
some statements on your sites about the user experience or point people to
where you have written it.  I find your input on this extremely helpfull and
would like to see more of this for modeling purposes if for no other reason.
What we really need and I have not seen is a comprehensive review along the
lines of what you have done earlier in this thread of A T so that we can all
get a good sense of what to expect and how to handle it.  I have seen
reviews of A T, but not to the extent that you took it in your set of
examples.  Mostly I've seen comparisons of one at over another and that in
and of its self is not bad, although it would be helpfull to see stand alone
reviews.  Looking at:
http://www.w3.org/wai/ua/ may provide some material and guidance here.  If
someone wants to review lynx with emax speak or some other combination, it
is welcomed and encouraged.


----- Original Message ----- 
From: "Ian Anderson" <lists@zstudio.co.uk>
To: "David Poehlman" <poehlman1@comcast.net>
Cc: <w3c-wai-ig@w3.org>
Sent: Friday, April 16, 2004 7:25 AM
Subject: Re: Screen readers - usage stats?




> There is no clear answer from here on your issue but perhaps there
is a best
> practices somewhere.  I always look for something independant of the
at
> issues to work with becaused you are on more solid ground if you do.

I agree completely. I look to WCAG for that solid ground. My own
experience helps me to apply it, and testing helps to validate common
use cases.

> Otherwise,to be fair, you have to do a lot more testing with other
environments
> as has been pointed out.  Even if market stats claim 95 percent of
screen
> reader sales are jaws, that may not even nearly accurately reflect
what is
> really going on around the world.

True. I think the UK AT market is different from the US one in many
ways.

>You can do one of two things with regard
> to market share.  You can do your best to find out what that market
is for a
> limitted audience say the uk for instance or you can assume that
your site
> needs to work with the most challenging environment that it should
> reasonably be expected to work with.

I try to do both. I can't test in everything, there isn't time. Coding
to valid HTML, and following WCAG should give a good chance for every
AT out there. If it doesn't, then there is arguably something wrong
with the AT, not my site.

My central dilemmas are about conflicts in the quality of user
experience that arise from testing even in a few cases. You can
imagine the geometric multiplication of conflicts that would arise in
testing twenty ATs.

I do not believe in lowest common denominator as a design principle. I
believe in graceful degradation, and in following generic design
principles to ensure access for all.

There's a sliding scale of user experience, though, and this is where
the optimisation comes in. The contentious issue seems to be who to
optimise for and why.

Received on Friday, 16 April 2004 09:06:19 UTC