W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 1999

Re: A few thoughts on using dynamic web pages to improve

From: Leonard R. Kasday <kasday@acm.org>
Date: Mon, 15 Nov 1999 15:57:53 -0500
Message-Id: <>
To: Scott Luebking <phoenixl@netcom.com>, phoenixl@netcom.com, w3c-wai-ig@w3.org
Hi Scott,

I agree that you can get optimal accessibility by generating pages
dynamically, tailored to the needs of the particular user.

As for whether I'm implying that there's a "primary" page... well, sort of,
but I prefer to word it differently, since that word might lead people to
think that it's somehow superior to the other pages, when all I mean is
that it's the page presented to the average non-disabled user who just
takes the default presentation.  So I'll call it the default page.

To be really simple and concrete, would the default page have ALT text?  I
think it should for a number of reasons:

1. If a blind and sighted user are working--or playing--together, it's
useful to have the same prsentation.  Lets say for example that the sighted
person has a question about how to use a screen.  If the blind user is
seeing a dropdown list and the sighted person is seeing radio buttons there
will be confusion. This sort of thing turned out very useful in
workstations I designed back when I was at AT&T.  As another example, a
parent who is blind may want to be looking at or talking about a screen
with a child who is sighted.  Sharing the experience more fully requires
access to everything on the screen, including "decorative" images.  This
last example is one a blind individual gave me back a few years ago when I
was asking about ALT text on decorative images.

2. Blind users can spot check that pages are equivalent and thereby gain
confidence that all are.  Or find bugs to report back.  It's a way for
users to do quality control.

3. Even if dynamically generated pages are really equivalent, they can
"send the message" that so called "text-only" pages are needed... and
promote their use where they are hand written and subject to error.

4. Accessbility features have advantages for non-disabled people.  E.g.
simple ALT text makes a page usable even if there are delays downloading.
A person would not want to switch to another page if they find images
coming in slowly.

So that's why I think that even if there are optimal dynamically generated
pages, the default pages should also be accessible.  I'm hoping you feel
the same about that.


At 10:06 AM 11/15/99 -0800, Scott Luebking wrote:
>Hi, Len
>Just to give a little perspective on how I see the issues.  I have to
>admit I don't believe in universal design in all situations.  For
>example, I don't want a universally designed wheelchair.  I want one
>that fits my needs.  My needs might conflict with another disabled
>person's needs.  For example, I need a wheelchair a certain height.
>Other people need their chairs other heights.  The approach that is
>often taken is to design the wheelchair so it can be configured to meet
>a particular person's needs.
>Basically universal design is appropriate when it is cost prohibitive to
>have many variations to accomodate each person's needs.  For example,
>having hamburgers which more closely meet each customer's preference is
>probably not that much more expensive than "universally-designed"
>hamburgers.  On the other hand, highly customer-specific cars can be
>very expensive.
>I believe that the web is and will be heading towards more
>personalization.  For example, according to this morning's San Francisco
>Chronicle The theme of Bill Gate's opening speech at Comdex was the
>"personal web".  In another example, I've developed some very simple
>free software which lets web sites generate web pages according to user
>needs.  I've been getting requests for the software from universities,
>portal developers, catalog developers, long distance learning,
>realty sites, etc.  The idea is that the web should adapt to the user rather
>than the other way around.
>Coming from that view, I thought it would be helpful to create my example
>of a web page tailored to meet a blind user's needs.  My example isn't
>complete in certain ways, but I wanted to get to the central
>A goal I had for the web page was to let a blind user be as efficient
>as possible.  A rough measure of efficiency was number of keystrokes
>ncluding any to the web page or to the access technology.  One way
>to reduce the number of keystrokes is to reduce the dependency on
>location to get information or action.
>Using an alert box is much more efficient than an index at the bottom of
>the page.  A good analysis of efficiency would include looking at
>the efficiency of getting the index information any time during
>the interaction with the page.  For example, if the user is in the middle
>of the page, how many keystrokes are needed to find the index at the
>end of the page.  Then, how many keystrokes are needed to return
>back to the exact position in the middle of the page.
>Breaking a web page into categories is much easier for a blind
>user to decide what to look at.  A sighted person can scan over a page
>which a blind person can't.  It is hard for a blind person to search
>a page based on concept rather than eaxct words.  Using the categories
>can help with that.
>Try listening to both pages in the example using a screen reader
>and not looking at the page.  See which is easier to naviage or interact
>The question you asked implies that there is perhaps a primary
>and a secondary form of the pages.  In a system of dynamically generated
>pages, the idea would be the match between the page and the user.
>While there are aspects of accessibility which should be included in
>the version of the page for sighted users in order to accomodate
>people with other disabilities, there might not be the same need for
>features needed for blind users.  For example, the issues of tables
>is a blind issue which is in conflict with visual appearance.
>Since it is pretty clear, that current CSS is not supported
>in a standard way by newer browsers or even supported by older browsers,
>it probably makes sense to let the pages for sighted users use
>tables for visual layout and effects, but limit tables in the blind
>versions to just data presentation when needed.
>In an environment where pages are being generated dynamically any way, I
>believe it would be easier to be able to generate multiple formats.
>For example, the web page formatted for a blind user is actually
>pretty easier to create.  It's almost pure text with forms.  No colors.
>No fonts.  No images.  No needing to line everything up to create
>a balanced, non-jagged presentation.  It's really easy to crank it
>out.  I believe adding post-processing will just complicate things.
>Post-processing is more apprpriate for environments where there is
>less control of how the pages are created originally.
>The javascript that is in the page now supports the event handling
>for both explorer and netscape.  Also, the architecture for the
>event handling is table driven so that adding events means
>just adding key event entries to the table.  The underlying code
>is kept quite independent from the initialization function.

Leonard R. Kasday, Ph.D.
Institute on Disabilities/UAP, and
Department of Electrical Engineering
Temple University

Ritter Hall Annex, Room 423, Philadelphia, PA 19122
(215) 204-2247 (voice)
(800) 750-7428 (TTY)
Received on Monday, 15 November 1999 15:56:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:13:45 GMT